Session S01: Poster Session & Refreshment Break IV (3:22 - 4:10 p.m.)

Bayesian reduction method on high-dimensional nonlinear models using deep probabilistic time series neural networks

Real-world problems are not entirely deterministic in the sense of the constraints imposed to them by data acquisition devices such as sensors. The available data obtained from sensors are limited, biased and possibly from multiple online sources. A way to learn the uncertainties through these complicated pipelines is crucial for a reduced representation of a complex system that captures the uncertainties. Fortunately, due to recent innovations in the software community, we have access to scalable Bayesian deep learning technology that can take advantage of such data to learn the uncertainties. It also naturally integrates with well-established neural networks and uses the same underlying framework to create a unified system.

To showcase Bayesian deep learning paradigm in fluid dynamics:

• we chose two nonlinear large scale benchmark models to generate full-order snapshots

• perform a classic reduction method to get a limited number of latent space time-series

• train two types of multivariate Bayesian time series models, Local Global Trend (LGT) and Damped Local Trend (DLT)

• compare diagnostics   

Optimizing computational fluid dynamic models to reduce simulation runtime

Optimizing the runtime in computational modeling of systems can save valuable time and resources. Long runtimes lead to significantly higher mistake costs, and limit how involved undergraduate students can be in research. This work has investigated the effect of variable mesh element sizes and simulation space size on the runtime and accuracy of in computational models, and found that both can have significant impact. The work has investigated how large the mesh element sizes can be while still producing accurate results. The larger mesh sizes in a model lead to a fewer number of elements, which lead to shorter runtimes. Additionally, we investigated how large a simulation space needs to be in order to produce accurate results as well. Again, minimizing the simulation space can also lead to fewer computational data points which leads to shorter runtimes as well. Here we report the impact of these types of factors on a simple finite element method computational fluid dynamics simulation.   

Dynamic mode decomposition with core sketch

With the increase in collected data volumes, there is an ever-growing need to develop computationally efficient tools to process these data sets. Modal analysis techniques have gained significant interest due to their ability to identify patterns in the data. Dynamic mode decomposition (DMD) relies on elements of Koopman approximation theory to compute a set of modes, each associated with a fixed oscillation frequency and decay/growth rate. Extracting these details from large data sets can be computationally expensive due to the need to implement singular value decomposition of the input data matrix. Sketching algorithms have become popular in numerical linear algebra where statistical theoretic approaches are utilized to reduce the cost of major operations. We put forth an efficient DMD framework, SketchyDMD, based on a core sketching algorithm that captures information about the range and co-range of input data. The proposed sketching-based framework can accelerate various portions of the DMD routines, compared to classical methods. The shallow water equations data is used as a prototypical model in the context of geophysical flows. We show that the proposed SketchyDMD is superior to existing randomized DMD methods based on capturing only the range of the input data.

    

Estimation of surface viscous stress from wave profiles using deep neural networks

The air-sea momentum and scalar exchanges are contingent on small-scale interfacial dynamics, which are crucial for climate and weather forecasting, significantly impacting many aspects of human life. To improve the predictive abilities of numerical models, it is essential to understand the behavior of wind stress, i.e., the sum of skin friction and form drag, at the ocean surface. Although skin friction contributes considerably to the total surface stress up to moderate wind speeds, it is notoriously challenging to measure and/or predict using classical physics-based numerical simulations.

Here, we present a supervised machine learning model to estimate the skin-friction drag of wind waves only from wave profiles and 10 m wind speeds, which are relatively easy to acquire. The input-output pairs are high-resolution wave profiles and their corresponding surface viscous stresses collected from experiments. The model consists of several convolutional neural network blocks with non-linear activation functions. Results show that the model can accurately predict the overall distribution of viscous stress; it captures the peak of viscous stress at/near the crest and its dramatic drop to almost null just past the crest, which is a robust indicator of airflow separation.   

A combined machine learning and data assimilation framework to model geophysical flows

The non-intrusive surrogate models, built on data acquired from sensors and satellites, are computationally inexpensive to model atmospheric and oceanic flows in comparison with numerical simulations, which makes them attractive for online deployments. Despite the recent success of data-driven prediction, online deployment of data-driven forecasting models might result in inaccurate predictions due to their biases regarding initializations, model architectures, and hyperparameters. The purpose of this study is to combine equation-free machine learning models responsible for predicting future states in the proper orthogonal decomposition (POD) latent space with the deterministic ensemble Kalman filter (DEnKF) to remove the biases and instabilities of the machine learning models. To this end, POD identifies dominant structures, and a long short-term memory (LSTM) technology predicts the dynamics of the system. The DEnKF algorithm corrects the prediction of the LSTM ensemble models in the latent space by incorporating noisy and sparse observations obtained from sensors that are optimally located with the QR pivoting method. The benefits of the proposed framework are successfully demonstrated by applying it to the NOAA Optimum Interpolation Sea Surface Temperature (SST) V2 dataset.   

Modeling of gas-liquid two-phase flows in a natural rock fracture-Application to carbon dioxide sequestration

Geologically sequestration of carbon dioxide (CO2) in brine-filled, subsurface formations has emerged as an effective approach to mitigate climate change. Within the low permeability subsurface rocks, there are fractures that act as natural fluid conduits. Therefore, understanding how CO2 moves when injected into an initially saturated rock fracture is critical for predicting carbon dioxide transport within fractured rocks.

Our study examined gas-liquid flow through models of rock fractures using a computational modeling approach. First, the Brown method was used to numerically generate a realistic three-dimensional rock fracture geometry with random roughness. Next, the fracture geometry was converted into a numerical mesh for gas-liquid flow calculations using the finite-volume solver ANSYS-FLUENT and the volume-of-fluid (VOF) method.

The simulation results for the case of a constant rate injection of CO2 into the initially water-saturated fracture were evaluated and presented for a range of conditions. The invading gas moved quickly, covering the large-aperture regions of the fracture. The effects of injection rate and interfacial surface tension between gas and liquid were studied. Relative permeability curves were developed to describe the gas-liquid flows in fracture. These permeability curves can be used in reservoir-scale discrete fracture models to predict CO2 motion within fractured geological formations.

    

Wake Induced Mixing in a Stably Stratified Environment

Stably stratified flows are encountered in different natural and industrial settings. The current research quantifies the effect three-dimensional wakes from a cylinder have on mixing downstream in a unidirectional stably stratified flow. The flows are simulated using open source spectral-element based high-order Navier-Stokes solver Nek5000. The direction of the flow is orthogonal to the direction of stratification, with stratification being out of plane with the cylinder. The effect of stratification on the drag, lift, and Strouhal number is quantified. Additionally, the effect of increasing stratification on the wake is quantified, including how the stratification changes the coherent structures downstream of the cylinder.   

Investigating the role of abyssal stratification in the propagation of vorticity throughout the water column: a Mediterranean example (Ionian Sea)

Analysis of in situ data from the Ionian Sea revealed the presence of variability with tidal periodicity in the deep layer, suggesting that the deep variability is actually related to the whole water column. The observations were made from 1999 to 2003, during the years of adjustment to the Eastern Mediterranean Transient, a major climate event that occurred at the end of the 80s. During the observation period, the Ionian deep layers were characterized by the presence of a stable water mass, the Ionian Abyssal Water, whose presence could be a key condition for catching such variability in the deep and for studying the role of the stratification on the propagation of the perturbation throughout the water column. The observed mean structure of the stratification suggests that a 4-layer scheme should be sufficient to have a realistic yet simple representation. To study how much and under which conditions a vorticity input can propagate, a quasi-geostrophic equation has been considered, with 4 coupled layers of arbitrary thickness and density, simulated with a custom-designed algorithm. This case study aims to give more insight into how energy stored by the deep layers can be released along the entire water column, contributing to the climate variability of the Mediterranean Sea.   

Some Insight into Frost Growth in Turbulent Moist Air Flow using Direct Numerical Simulations

We developed a new model to predict frost growth over a flat plate maintained at subfreezing temperatures and subjected to humid turbulent air flow. The model employs a dynamically coupled air-frost interface. The air phase has been modeled using direct numerical simulations, while the frost phase has been modeled from first principles using the conservation equations of mass and energy. Coupling of the two phases is done using either the immersed boundary method or by deforming the bottom boundary and using a body-fitted grid. Due to the vastly different time scales between the fast developing turbulent flow and the much slower frost phase, a slow-time acceleration technique has been implemented to make the simulations feasible by accelerating the frost growth process. The model has been validated against laboratory experiments and then used to predict frost growth under a variety of free-stream and plate conditions. We observed that the Nusselt and Sherwood numbers could be properly scaled so as to become primarily dependent only on the Reynolds, Prandtl, and Schmidt numbers. A series of simulations covering a range of shear Reynolds numbers between 100 and 2000 were then used to extract the Nusselt and Sherwood number dependencies on the Reynolds number.   

Author not Attending

Mesoscale convective systems (MCSs) are thunderstorms up to hundreds of kilometres in size which can have devastating impacts. They dominate the weather systems in West Africa, where they contribute up to 90% of the rainfall in the region during boreal summer, but currently it is very difficult to forecast when they will initiate and where. New developments in regional convection-permitting models have improved representation of these systems, with more realistic distributions of MCSs with more comparable sizes and speeds to those observed using satellites. Such models are an excellent tool to explore the dynamics of MCSs in more detail.

Previous studies have indicated that MCSs over West Africa can be triggered by synoptic and local convergence fields, and studies of MCSs on other continents have shown that new storms can be initiated by vortices that remain after existing MCSs have dissipated. Using CP4-Africa, a convection-permitting regional climate model, and a dataset of tracked storms in the simulation, we investigate what dynamical factors are most strongly associated with MCS initiation. Determining the dynamical environment in which storms initiate could be crucial to improve forecasting of such systems in operational models.

    

A numerical analysis study on lithium-ion battery thermal management system using cold plate

Nowadays, lithium-ion battery is widely used in Electric Vehicle because of its high-capacity and high-voltage. But there are still some drawbacks about batteries stability. The aim of our research is to find ideal cold plate options which differed the width and number of channels. In order to estimate the temperature distribution and heat transfer rate, MSMD – Newman P2D model in Ansys Fluent program is used. Prior to compare the heat transfer rate of battery surface temperature using differ cold plates, we calculate the surface temperature of battery (LiFePO4) at different discharge rates at 2C, 3C and 4C to explore the batteries characteristics. After that, we attach the two cold plates to batteries both sides (front and back) and estimate the heat transfer rate of battery surface which contacted with cold plate, the pressure drops between inlet and outlet of channels during discharge process. Also, we calculate j factor and f factor that appropriate for estimating cooling performance of cold plate. To confirm the most efficient cold plate options, trade-off between the heat transfer coefficient and the pressure drop is important also the relations between two factors (j factor, f factor).   

Solar Desalination Chimneys

In this study, a design for a solar desalination system comprised of a solar chimney and humidification-dehumidification system is proposed. In a solar desalination chimney, solar energy creates a buoyant hot air flow that passes through the entire system. The wind turbine inside the chimney generates mechanical or electrical energy for pumping water into the desalination unit without additional energy. To investigate the potential of the proposed system, validated one-dimensional models for solar chimney and direct contact packed-bed humidification-dehumidification are used. Three solar chimney configurations are examined for their desalination potential. A solar desalination chimney might be incapable of working due to the pressure drop inside the desalination unit. For Florida and Kerman solar chimneys, results indicate that in regions with abundant solar irradiation autonomous desalination is possible. Desalination analysis shows that a large-scale solar chimney such as Manzanares with modification is capable of providing freshwater for about 800 households; while a small-scale such as Florida with a surface footprint of 66 m² could satisfy the freshwater needs of a household.

    

Numerical investigation of the effects of wing bristles and wing flexibility on the forward flapping aerodynamics of the smallest flying insects

We used computational fluid dynamics to investigate the effects of bristles and wing flexibility on the forward flapping aerodynamics of the smallest flying insects. The immersed boundary method was used to solve the fully coupled fluid-structure interaction problem of a pair of flexible wings immersed in a three-dimensional viscous fluid. To determine the effects of bristles, three wings were considered that ranged from least to most bristled. The results suggest that at Re relevant to small insect flight, rigid-bristled wings generate nearly as much average vertical force and thrust as rigid-solid wings while providing the benefit of lower wing mass. To investigate the effects of wing flexibility, four flexible-solid wings with spanwise flexibility were considered and a new parameter called deflection angle was defined to characterize their flexibility. Based on the results, it appears that at Re pertinent to tiny insect flight, adding a high degree of spanwise flexibility to rigid-solid wings deteriorates their average vertical force and thrust to a large extent. For moderately spanwise-flexible wings, the results suggest that they generate nearly as much average vertical force as their rigid counterparts while providing the benefit of increased average thrust.

    

Computational investigation of disease-conducive flow conditions in the venous valve sinus

While venous valves are critical to proper function of the circulatory system, they are also the most common locations of origination of venous thromboembolism (VTE). VTE, which comprises deep vein thrombosis (DVT) and pulmonary embolism (PE), is a leading cause of death in the United States, especially for those who are immobile for extended periods of time (such as for hospitalization or long haul plane rides). Here, we use a three-dimensional (3-D) fully-coupled fluid-solid interaction (FSI) model to investigate the effect of venous valve 3-D shape and stiffness on flow conditions. Specifically, we consider flow conditions such as fluid stasis, high residence time, and low shear stress, which are thought to be important in thrombus formation. Our findings will help in increasing understanding of this disease, and represent an important step in determining how to better identify hospital patients at high risk for VTE, which is highly preventable if appropriate measures are taken.   

Integration of Catheter Based Hemodynamic Data with 3D Rotational Angiography for Computational Hemodynamics Modeling of Congenital Heart Disease.

3D Rotational Angiography (3DRA) is standard-of-care for congenital heart disease (CHD). However, 3DRA is not commonly the imaging of choice for integration with CFD analysis, unlike Computed Tomography or Magnetic Resonance imaging. 3DRA Imaging has many advantages over other imaging modalities such as higher resolution, no additional increase in radiation, and availability of active hemodynamic information. These can be advantageous when paired with CFD to enable improved CHD treatment by providing physicians visualizations of how blood flows throughout the vasculature of interest pre- and post-surgery. Physiologically consistent flow visualization requires systematic integration of catheter hemodynamic data with 3DRA imaging. In this study, we present 3D in silico cardiovascular CFD modeling approach to demonstrate the use of 3DRA imaging paired with catheter derived hemodynamic data. Simulation case studies will be discussed for four cases involving different anatomies, physiologies, and flow states. Flow ratios and distributions based on simulation results will be used to illustrate how closely these results match with the hemodynamic information assigned for each model from clinical measurements.   

Coherence at many scales in the vocal tract and jet during speaking

The intricate geometry of an anatomically accurate vocal tract and variations in flow rate develop a wide range of turbulent scales both in the tract and the external jet during breathing and speaking. The "train of puffs" or turbulent pulsed jet complicates the dispersion of aerosolized particles emitted from the mouth during such events. We study the evolution of coherent structures of steady and pulsed jets during speaking using geometrically-resolving large-eddy simulation with the curvilinear immersed boundary method capturing the internal vocal flow and the external speaking jet characteristics of an anatomically accurate vocal tract geometry. Simulations reveal complex coherent structures such as attached and shed vortices and recirculation zones over a wide range. Large vortical features are observed in the glottal region and evolve and break up through the pharynx and oral cavity. Spectral analysis and proper orthogonal decomposition are employed to quantify the dominant coherent structures. It is demonstrated that the intricacies of geometry create high vorticity zones and coherent structures.   

Uncovering the Impact of Morphological Features on Vocal Fold Function

The human phonation involves complex interactions of airflow through the larynx and the flow-induced vibrations of the vocal folds. The vocal fold of men and women differ in their anatomic and physiological characteristics, which have repercussions on their vibratory patterns. However, no previous research has attempted to explore the implication of differences in the structural dimensions on phonation process. The objective of this study is to perform a parametric analysis to explore the effect of variation in these morphological features on the glottal aerodynamic and the tissue deformation in a systematic manner. Geometric variations studied include vocal fold thickness and length, as well as the vocal folds depth. The fluid-structure interaction simulations of the dynamic vocal folds coupled to the unsteady, turbulent motion of the air past them is performed for nine cases with variable dimensions and vocal fold vibration and the resulting model outcomes such as pulsatile glottal jet, tissue deformation and several metrics for stress distributions are analyzed. We observed that the vocal fold thickness and depth are the parameters that influence the tissue deformations and stresses during phonation, while there was little effect from the vocal fold length.   

Choanoflagellate Model Using Regularized Stokeslets and Segments

Choanoflagellates are the closest living relative to animals and act as both unicellular and colonial eukaryotes. Their morphology consists of an oblong cell body with a flagellum for movement and a collar composed of microvilli that are used to capture bacteria. To simulate their feeding behaviors, our current computer simulations calculate the inward flux of fluid to the collar and the choanoflagellate translational and rotational velocities given its prescribed flagellar movement. However, the method of regularized Stokeslets (MRS) requires using thousands of points to discretize a morphology of a single cell to match the simulated results with experimental data. Since a large portion of the discretized points is on the collar, the method of regularized segments (MRSE) can  simplify the model. The idea is, along each microvillus, replacing a cluster of discretized MRS points with a regularized segment that has linearly distributed forces. In this project, we found an optimal number of regularized segments to drastically reduce total memory consumption and computation time. The simulated results from the mixed method of MRS and MRSE matched the experimental velocity data and the inward flux calculations of the original MRS model. We then extended this simplified model from one cell to two cells to quantify how their angles affect the outputs. With the reduced computational strain and memory usage due to the mixed method, we were able to study the feeding patterns of larger colonies that were impossible to simulate using the MRS method. In the future, we will be able to simulate various colony shapes to test the hypothesis that multicellular life forms do have advantageous feeding patterns. 

    

Turbulent Flow of Polymer Solutions in a Square Duct Roughened with Transverse Ribs

In this research, a direct numerical simulation of the turbulent flow of a polymer solution is performed in a ribbed square duct of different blockage ratios, with the FENE-P model used to simulate the presence of polymers. The effects of parameters defining the viscoelastic behavior of the polymer flow such as the Weissenberg number are investigated for both cases with and without riblets. The numerical results are compared with those of a smooth duct polymer flow as well as with those of a Newtonian ribbed duct flow. The results indicated that the turbulence field is influenced by the rib elements, the four duct sidewalls and also the polymer additives. It was observed that, in the Newtonian ribbed duct flow, when increasing the blockage ratio, due to the strong shear layer formed immediately above the rib crest, strong turbulent secondary flow motions are formed in that region, which facilitates the transport of Turbulent Kinetic Energy (TKE) in the cross-stream directions. In the polymer ribbed duct flow, the aforementioned strong secondary flow is altered, so that the locations of the maximum vorticity move away from the wall, towards the centerline, and also the circulation is improved in each of the 8 sectors of the duct.

    

The effect of viscoelastic flow on instabilities of plane Poiseuille problem with porous walls

The primary aim of this study is to investigate the linear stability of pressure-driven viscoelastic flow in a channel with porous walls at the top and bottom. Unlike the Newtonian flows, the polymer solutions create destabilizing effects in streamwise perturbations at different Reynolds numbers in smooth channels. The porous media in the same system, on other hand, also have a significant impact on destabilizing the flow. Here, we combined these two destabilizers and studied different types of unstable modes analytically and numerically by varying the dimensionless parameters that govern the flow stability (Reynolds number, permeability parameter α=H/κ^1/2, elasticity number E=λμ/(ρL² ), and the ratio of solvent to solution viscosity β=μ_s/μ, while the depth ratio is constant; here, ρ is the fluid density, μ is the fluid viscosity, λ is the relaxation time, and L is the channel half-width). Also, using the linear stability analysis for a wide range of Reynolds number and wavenumber indicates that due to the elasticity, the critical condition and stability behavior of channel flow with porous walls change considerably. 

    

Linear Stability Analysis of a Particle Laden Poiseuille Couette Flow Over a Porous Layer

In this study, we discuss the linear stability analysis of a particle laden flow in a Poiseuille Couette flow configuration with a bottom Brinkman porous layer and an impermeable top moving wall.  The Dusty-Gas Model is used to model the flow as a mixture of pure fluid and particles due to the significant impact of the particles. We model the flow in the porous layer by using the Brinkman equation and couple it with the Navier-Stokes equation used to model the flow in the fluid layer. The effects of the Couette flow component, porous to fluid layer depth ratio, permeability, porosity, particle concentration and more on the linear stability analysis are discussed. The results show that changing the listed components effect the transition between stabilized and turbulent flows.

    

Atwood number effects on the isothermally stratified compressible Rayleigh-Taylor instability

The coupled effects of the variable-density and background stratification strength on the growth of the fully compressible 2D single-mode two-fluids Rayleigh-Taylor instability (RTI) are examined using direct numerical simulations (DNS) with varying Atwood and background isothermal Mach numbers. Compared to small Mach number (weakly stratified) case, at larger Mach number (strongly stratified) cases, we observe more asymmetric growth rates between the bubble side, regions with light fluid penetrating into heavy fluid, and spike side, regions with heavy fluid penetrating to the light fluid, for increasing Atwood number. This finding suggests that the asymmetric growth of the RTI mixing layer is enhanced by the compressibility and variable-density effects become observable at relatively small Atwood number under strong background stratification. We also analyze the combined Atwood and Mach number effects on the vortical dynamics and mixing of 2D compressible RTI.

    

A CFD Model of Evaporation in Liquid Hydrogen without the Need for Tuning Coefficients

The Hertz-Knudsen-Schrage equation, derived from kinetic theory, describes evaporation and condensation but requires accommodation coefficients as inputs. Reported values of the coefficient are controversial and span 3 orders of magnitude for common fluids such as water. The data for cryogenic fluids is severely limited. Computational modeling of evaporation in liquid hydrogen is critical to evaluate the cryo-storage stability of fuel depots to enable long-term space missions. However, the coefficient is often reduced to a tuning parameter to achieve numerical stability. Recent results indicate that transition state theory could provide an analytical description of the accommodation coefficients based on the physical parameters of the liquid and the vapor. Here, a new computational method to model evaporation is developed using a combination of transition state theory and kinetic theory to alleviate the need for tuning coefficient values. A custom CFD model for steady evaporation from a liquid hydrogen meniscus is built using user-defined functions in Ansys Fluent. Numerical simulations are conducted using inputs from an experimental cryo-neutron imaging dataset. At each instantaneous time step, a sharp liquid-vapor interface is assumed, held static and a zone of cells on either side of the interface is identified as an “active region”. This region is instrumented with custom non-uniform mass and energy sources/sinks. The accommodation coefficient is computed in-situ by probing corresponding liquid and vapor cells on either side of the interface. The non-uniform evaporation flux is then integrated over the meniscus to obtain an instantaneous net evaporation rate and compared with the experimentally measured evaporation rate for validation.   

Positivity-preserving numerical method for thin liquids film on vertical cylindrical fibers.

When a thin liquid film flows down on a vertical fiber, one can observe the complex and captivating interfacial dynamics of an unsteady flow. Such dynamics are used in various applications due to their high surface area to volume ratio. Recent experiments indicate that when the flow undergoes regime transitions, the magnitude of the film thickness changes dramatically making it difficult to develop a numerical method accounting for the changes.  We present a computationally efficient numerical method that can maintain the positivity of the film thickness as well as conserve the volume of the fluid under the coarse mesh setting. 

    

Drag on Prolate Spheroids for a Wide Range of Mach and Reynolds Numbers at Zero Angle-of-Attack

The aim of this poster is to showcase ongoing efforts regarding the characterization and correlation for the drag coefficient of an isolated spheroid. The effort attempts to expand on previous efforts that focus on low Mach, where in this effort the input range is expanded to a wider range of Mach numbers from subsonic, to low-hypersonic flow (i.e., 0-6). The effort also spans Reynolds numbers relevant to particulate flow (0 to 250) and aspect ratios of 1 to 4. In addition, the effort only considers the effect of the spheroid at a zero degree angle-of-attack, hence, aims to provide insight on one of many geometric sensitivities. The effort uses computational fluid dynamics (CFD) and simulates the flow over a stationary isolated spheroid using STAR CCM+, a commercial CFD package. The computed drag forces were then appropriately normalized to obtain the drag coefficient. Using these data, the effort developed new correlations for the drag coefficient relevant to varied Mach and Reynolds numbers with the added dimension of aspect ratio.   

Vortex Dynamics using the Principle of Least Action

Variational formulation of vortex dynamics has a long history with a very rich literature. The standard Hamiltonian that describes the dynamics of interacting constant-strength point vortices is the Kirchhoff-Routh (KR) function. This function was not derived from basic definitions in classical mechanics (it is not the system kinetic and potential energies summation). Rather, it was devised such that its hamiltonian dynamics match the already known first-order differential equations of motion for constant-strength point vortices given by the Bio-Savart law. While this approach is widely accepted, it is an ad-hoc one that does not allow for extension to, say, time-varying vortices. Here we develop a new variational formulation of vortex dynamics using the principle of least action, and the Lagrangian density is given by the pressure. A system of non-deforming free vortex patches of constant-strength is considered as a case study. Setting the first variation of the action integral with respect to the position coordinates of the vortices results, for the first time, into second-order ODEs, defining vortex acceleration not velocity. Interestingly, for constant-strength point vortices (the limit to vanishing core), the current formulation reduces to KR dynamics. However, for vortices with finite cores, the resulting dynamics are richer than that derived from KR. For example, a pair of equal-strength counter-rotating vortices starting with the same velocity, will continuously attract and repel each other in contrast to the constant-relative-distance motion predicted by the KR dynamics. And if they have different initial velocities, many interesting patterns may occur, which can't be even handled by the standard KR approach. Finally, the fact that this new variational formulation is derived from mechanics first principles, allows straightforward extension to arbitrary time-varying vortices.

    

Spacial characteristics of vortices in a channel flow – vortex generation with shear independence

The present study investigates the topological characteristics of vortices in the generation and development in terms of the local topology, with relating to the spacial or shear condition of a channel flow. The generation process of vortices can be followed in the pre-swirl coordinate system and vortex space coordinate system that are Galilei invariant associated with (predicted) swirl plane. Then directions of the swirl planes of vortices in the generation and after development are analyzed, relating them to the primary directions of the turbulent flow. The analysis clarifies that a vortex near the viscous sublayer is generated with a swirl plane parallel to the wall of the channel flow, which indicates that the intense shear in this region does not support in the vortex generation in terms of the local flow, or rather prevent it. A vortex in a flow region with less shear is generated with random direction that may be similar to the vortex generation in homogeneous turbulence. On the other hand, if a vortex is developed, an effect of shear on the direction of the swirl plane decreases. These results indicate that the shear does not always support the vortex generation and that the contribution of the generation depends on its characteristics.   

The passing feature of the bundles of the vortical axes, and topological relationships between the vortex and the vortical axes in an isotropic homogeneous turbulence

The present study analyses the characteristics of the vorticity lines, the eigen-vortical-axis lines, and the pressure minimum lines in the vortical regions, and analyses the topological relationships in the three lines. This analysis applies the physical quantity called swirlity φ that evaluates the unidirectionality and intensity of azimuthal components of the subjected vector field. It also applies the theory of local axis geometry to analyse the topological characteristics of the above three lines in the vortical region. The results show that the vorticity lines and the pressure minimum lines have a swirling feature, while the eigen-vortical-axis lines do not tend to swirl in the vortical region. And, it shows that the directions of the rotation of the vorticity lines and the pressure minimum lines may be changed with the intensity of swirling along a vortical region. In addition, the angles between the two tangential vectors in the three lines show that the directions of the vorticity lines and the pressure minimum lines have a high correlation. Therefore, these two lines have a similar feature of the passage in the vortical region.   

An investigation of application of sourcity in identification of a quantum vortex: identification of divergence characteristics of probability flow

The present study investigates the application of a geometrical quantity, sourcity, in the identification of a quantum vortex in a quantum turbulence. This quantity specifies the unidirectionality and intensity of the radial flow in a vector field in terms of the local flow geometry, which has been used to investigate the detail geometry of vortices in a classical turbulent flow. A quantum vortex is difficult to be identified with focusing on a swirling flow feature because the gradient tensor of superfluid has no rotational component except points of topological defect, i.e., quantum vortices. On the other hand, probability flow or the gradient vector of absolute value of a wave function has an outflow characteristic in a quantum vortex, which suggests that the application of the sourcity may contribute to the identification of this feature. A joint probability density function of the sourcity and magnitude of a wave function indicates a high correlation in a Direct Numerical Simulation of the Gross-Pitaevskii equation. The sourcity may be a candidate in the identification method of a quantum vortex.   

Computational Study of Jet-Actuation Design for Supersonic Flow Control

An active flow control technique of micro jet-array is applied to a Multi-Aperture Rectangular Single Expansion Ramp Nozzle. The nozzle configuration introduces a bypass stream to the primary stream with a splitter plate to separate the flows. This study investigates the effects on the high-speed jet flow of an active flow control strategy by introducing a micro jet-array blowing into the region where the two streams coalesce using both experiments and simulations. To aid the experimental effort in implementing the actuators, we perform simulations of the interaction between the micro jet-array and the cross-jet flow to characterize the actuation profile at the outlet of the micro-jet array. Multiple simulations are conducted at various nozzle pressure ratios. The simulated results will be compared with the pressure data collected from experiments through near-field pressure transducers at various locations on a deck plate. Meanwhile, time-resolved schlieren images are used to visualize the flow behavior over the deck plate. In this joint effort, the experimental measurement will validate the computational results, and the simulated result will provide additional information in the active flow control study investigated in the experiments.   

Dynamic Stability of the Mars Science Laboratory Entry Vehicle in Supersonic Flow

The dynamic stability of a blunt body vehicle is complex and plays a significant role in an atmospheric entry. This study investigates the underlying flow physic of the Mars Science Laboratory (MSL) entry capsule at the supersonic flow regime. Initially, a two-dimensional geometry of the entry vehicle is created to perform a static computational fluid dynamics simulation. Then, the entry vehicle is simulated dynamically, where translations and rotations of the vehicle due to the forces and moments are allowed. The unsteady Reynolds averaged Navier-Stokes equations (RANS) are solved for the Mach numbers ranging from 2 to 3, and results are analyzed using the reduced-order model (ROM) and dynamic mode decomposition (DMD). The relationship between the variation of Mach numbers, pressure coefficients, and stability parameters will be presented. The computational analysis offers substantial knowledge to mitigate the risks associated with designing and probing the physical mechanisms behind the atmospheric entry vehicle dynamic stability.   

Efficient evaluation of the surface roughness effect on boundary layer using quasi-statically transforming roughness shape

Surface roughness has a significant influence on the boundary layer. Depending on the shape of the roughness, it sometimes triggers turbulent transition immediately and at other times laminarizes the flow at the downstream side. Although direct numerical simulation (DNS) is a powerful tool, a lot of computations are required for assessing the influence of one roughness shape on the boundary layer. In this study, we develop an efficient method to examine a series of roughness shapes in one simulation by changing a shape parameter quasi-statically with a volume penalization (VP) method, where we should sufficiently reduce unsteady numerical noise generated at the moving shape by adjusting the moving speed and the permeability of the VP method. The subsequent change of the disturbance is observed at a downstream position, where we allow for the delay time by assuming that the disturbance takes time to be advected to this position. We demonstrate this method for optimizing discrete and sinusoidal roughness elements placed on a three-dimensional boundary layer.   

A modular method for estimation of velocity and temperature profiles in high-speed boundary layers

The variation of fluid properties due to viscous heating affects the mean velocity and Reynolds stress profiles in compressible turbulent boundary layers, thereby also altering the resultant skin friction and wall heat transfer. In the current work, we present a method based on the work of Huang et al. [AIAA Journal 1993 31:9, pp. 1600-1604] that estimates the boundary layer velocity and temperature profiles, and therefore also the skin friction and heat transfer coefficients, for a given Mach number, Reynolds number, and wall thermal condition. The proposed method is modular in the sense that it works with multiple variable-property scaling formulas (or, "velocity transformations"), velocity-temperature relationships, viscosity-temperature relationships and equations-of-state. The current method has been shown to make predictions with improved accuracy as compared to the current state-of-the-art, the Van Driest II method.   

Author not Attending

Dynamic stability plays a vital role in the design and evaluation of atmospheric entry vehicles (AEV). Blunt body capsules are great for reducing the extreme temperatures of atmospheric entry but typically become dynamically unstable near the lower end of their trajectory. The dynamic instability causes the amplitude of the angle of attack to increase with time until the growth reaches an equilibrium point or the oscillation reaches a divergence point which sends the blunt body into a tumble. The dynamic stability of AEVs is mainly studied through ballistic range experiments and flight tests. Recent advances in computational fluid dynamics (CFD) and dynamic mode decomposition (DMD) provide a novel approach to predicting free flight dynamics, as most CFD simulations are static aerodynamics. This presentation will use static and dynamic CFD simulations to identify the stability derivatives associated with the Earth Entry Vehicle (EEV). Dynamic simulations are performed using a forced oscillation by a sliding mesh model. The time-resolved flow field of the EEV wake is investigated numerically by employing unsteady Reynolds-averaged Navier–Stokes (URANS) equations. The flow field snapshots will be used in a data reduction technique to reveal hidden details in the complex wake flow.   

Reynolds number dependency on the aerodynamic performance of corrugated wing in unsteady motion

The dragonfly wing, which has a corrugated structure, may cause higher aerodynamic performance than the flat wing at the low Reynolds number regime (Re \simeq O(10^3)). However, the details are unclear. Not many studies have investigated detailed lift generation process in unsteady motion of the corrugated wing, which is observed in nature. In this study, we concentrate on the translating motion starting impulsively from the rest state in two-dimensional space, which is one of the fundamental modes of unsteady motions.

The direct numerical simulations revealed that the lambda vortex, generated on the upper surface of the wing, is a key for aerodynamic performance. When it is stuck in the V-shaped structure of the corrugated wing, the performance was better.

In the presentation, we report the Reynolds number dependence on this mechanism. Lift enhancement associated with the lambda vortex behavior is more pronounced at the higher Reynolds numbers (1000<Re), whereas the simulations were performed in the range 100<Re<4000.   

Effect of Aluminum Particle Size on Solid Rocket Combustion

Small particles of metals such as aluminum (Al) are often chosen as the fuel in solid rocket propellants because of their high energy density. Experimental conclusions have shown promising results for the use of nano aluminum particles to increase burning rates. Computational Fluid Dynamic (CFD) studies validating these experimental findings are limited. A numerical simulation comparing conventional and nano size Al particles is conducted using the Euler-Lagrange method in the CFD software Star CCM+. The model uses a multicomponent reactive gas flow to simulate the essential combustion reactions and an evaporation model to simulate the Al droplet evaporation. The simulation confirms the correlation between small particle diameter size on an increase in burn rate and combustion characteristics.

    

Optimization of Polo Kayak Design for Drag Reduction

Pressure drag is a significant portion of the drag forces on small boats such as kayaks and rowing shells and is a result of both profile and wave drag. These drag components depend on the boat profile, i.e., frontal area and volume distribution. As the boat begins to plane the drag coefficient decreases sharply facilitating a further increase in speed. This is highly beneficial in the sport of kayak polo, where players seek to accelerate and move quickly on the court. This poster presents a parametric study of polo kayak geometry using Computational Fluid Dynamics (CFD) simulations to reduce drag across several Froude numbers. A Volume of Fluid (VOF) approach models the fluid, while a Crank-Nicolson rigid body motion solver with two degrees of freedom (heave and pitch) is employed to model the boat motion. The starting geometry is representative of current polo kayak designs with a characteristic length of 3 m and width of 0.53 m. This geometry is parametrized based on the cross-sectional shape at ten discrete locations along the kayak longitudinal axis. The distribution of the cross-sectional area is studied with respect to drag and lift forces to suggest a design that is optimized for drag reduction.

    

Correlation between severity of a modeled aortic valve stenosis and its acoustic spectrum

Aortic stenosis is one of the most common heart valve diseases that occurs due to the narrowing of the aortic valve. Heart murmurs are abnormal sounds caused by turbulent blood flow due to the narrowing of the valve. This study seeks to understand the relationship between stenosis severity and energy content at different frequencies. The experimental setup includes two distinct cases; a circular narrowing and a simplified semi triangular shape to mimic an opened aortic valve. Each shape is 3D printed with eight matched restrictions ranging from 0% to 82% blockage ratio. A pulsatile pump regulates the flow and a contact microphone collects the sound pressure levels at the narrowing for all conditions. Flow rates in all test cases have been set to match the Reynolds numbers through the opening which are based on healthy human heart conditions. A spectral analysis shows the energy content is influenced by both the stenosis severity and the Reynolds number of the flow. As the restrictions become increasingly narrow, the energy present at higher frequencies increases, most notably in the 300-500 Hz range. The same trend is present with increasing Reynolds number. This provides a general look into identifying characteristic frequencies that indicate the severity of a restriction in this modeled pulsatile flow, and highlights the interactions of parameters that increase the complexity of identifying stenosis.   

Development of an experimental slicone model of venous valves

Venous valves are important components of the circulatory system. These valves open and close with pressure oscillations due to contraction and relaxation of the surrounding skeletal muscle, allowing forward flow and blocking reverse flow, thus enabling return of blood from the lower extremities against gravity back to the heart. However, the venous valve region is also the origin of the majority of instances of venous thromboembolism (VTE). VTE, which includes both deep vein thrombosis (DVT) and pulmonary embolism (PE), is a leading cause of death in the United States and is an especially serious concern for those who experience extended periods of physical inactivity (such as hospitalization or long plane rides). Here, we describe a silicone model of the venous valve region, including a highly-flexible model valve. We demonstrate the ability of our model valve to open and close with pressure oscillations, and compare the resulting flow to that found in numerical simulations. This model allows us to control morphological conditions such as valve shape and stiffness to investigate their effect on the resulting flow. Further, this model will allow for validation of numerical results, and will give insights into the possible effects of imperfections and additional conditions not typically considered in numerical simulations.   

Permeability of the lateral air flow through pillar-like nanostructures for microfluidic applications

In many microfluidic devices represented as Lab-on-a-Chip, unwanted air can go inside during preparation processes. They can impede the flow of the device, and reduce the effectiveness of certain functions such as analysis and separation. To prevent this, we suggest a method for removing the air inside a channel using the nonwetting nanostructured substrate. In this concept, the permeability of the air passing between the nanostructures is very important. We prepared the pillar-like nanostructured substrate using the combination of dilute Ag-ink sintering and a metal-assisted chemical etching on a Si-wafer. The PDMS-based microchannel was fabricated and attached to the nanostructured substrate for experiments. By measuring the pressure change inside the initially pressurized microchannel, the permeability of air is compared with existing theories to analyze the results. Kozeny factors were determined, and the results were well explained with the previous studies. This work is to decide appropriate operating conditions for microfluidic applications, so the results are meaningful to suggest quantitative values for the appropriate operating pressure and expected degassing time in practical applications by evaluating the air permeability through the irregular nanostructures.   

Application of an accessible microfluidics platform to diffusion into dead-end pores

In this work, we use a rapid prototyping technique to produce inexpensive, flexible microchannel chips out of double-sided adhesive tape using a craft-cutter in seconds with high precision. We then visualize the passive diffusion of fluorescein dye solution using an OpenFlexure microscope, which is an open-source device centered around a Raspberry Pi housed in a 3D-printed shell. We adapt the microscope to our application using the picamera python package and a color correction calibration method to quantitatively track the diffusion of fluorescein dye into micropores over a prolonged period of time. Our measurements are compared to analytical solutions of an effective 1D Fick-Jacobs equation for the case of both rectangular and trapezoidal cross-sectional geometries with good agreement. Open questions and future applications of our accessible microfluidics platform will be discussed.   

Mechanism of secondary entrapment of bubble.

When a drop strikes the water's free surface, a cavity forms beneath the surface. During the collapse of this cavity, a bubble entrapment is observed and this is generally called a "secondary bubble entrapment" or "irregular entrapment," since bubble formation may or may not occur under the same impact conditions. The mechanism of this secondary bubble entrapment differs slightly from the mechanism of the primary bubble entrapment (regular entrapment) mechanism. This is due to an increase in the Weber number and Froude number, which increases the maximum cavity depth and changes the cavity shape. This bubble formation is only possible if the crown collapse causes wavefronts to travel down the cavity walls at the same speed. These wavefronts arrest the cavity retraction slightly above the maximum cavity depth, steepen the cavity walls, and transform the cavity base into a single stepped cone. When the wavefronts converge near the cavity's axis of symmetry, the single stepped cone base pinches off as a bubble and leaves the cavity as a truncated cone. The speed of the wavefronts traveling down the cavity walls would probably depend on the drop shape at the time of impact on the free surface of the water.   

Evaluation of splashing threshold of impacting droplets based on automated experiments

Splashing of impacting droplets is an important phenomenon which frequently happens both in nature and industry. Thus, many studies have been carried out to model the splashing threshold such as K-Parameter proposed by Mundo et al (1995) and the splashing threshold model based on aerodynamic lift force by Riboux et al. (2014).

However, the judgement of splashing threshold heavily relies on human eyes, which is subjected to human error. Moreover, due to the insufficient amount of data, statistical errors are likely to occur. Therefore, the purpose of our study is to automate a drop-impact experiment for obtaining a large amount of data. Also, we improve the proposed splashing threshold models into a universal model that is robust in various impact conditions by reducing these errors. The automated experimental system is built with with a robot arm, a high-speed camera, and actuators. This system can get a massive amount of high-quality image sequences of impacting droplets. The results obtained in this study is analyzed by AI for evaluating exisiting splashing thresholds.   

Evolution of the "First Bubble": Bubble and Jet Formation Within a Wine Bottle

When a liquid-filled bottle is inverted and allowed to empty, aside from the "glug" we hear, what do we see? What happens to the first bubble that rises into the neck? How does it form and evolve over time? We report on a set of experiments to address those questions, qualitatively and quantitatively, using a wine bottle. Due to the brief timespan of the first bubble evolution, a high-speed camera was needed to enable detailed observations of behavior. A MATLAB program was written to analyze these high-speed images and collect quantitative information. To explore the effects of liquid viscosity, we utilized a range of water-glycerin mixtures which yielded a wide range of Reynolds (10–20,000) and Weber (10–500) numbers, and limited range of Eötvös numbers (250-370). In our experiments we observed transitions in bubble characteristics (shape, size, motion, etc.) as liquid viscosity increased. After initial pinch-off within the bottleneck, bubbles in low viscosity liquids produced ejector jets which penetrated the top surface of the bubbles. Jet behavior varied with viscosity, exhibiting distinct changes in shape, pattern, and jet tip velocity. Supported by striking visuals, our findings provide deeper insight into the first bubble's formation and breakup as a function of viscosity.   

Reproducible ejection of highly-viscos liquid jets

We have developed a device to eject a highly-viscos liquid of 300-1,000 cSt as a focused jet using an impulsive force (Onuki et al. Phys. Rev. Applied 2018). However, variations in shape and area of the coated droplets were observed during periodic ejection of jets. Therefore, this study aims to elucidate the factors of the instability during periodic jetting and to improve the quality of coating. For that, we experimentally observed the jet behavior during periodic jetting. It is observed that the meniscus position and the shape of the meniscus changes with each ejection. Since the rise of the meniscus position can change the gradient of pressure impulse, which causes the acceleration of the fluid, and the meniscus shape can affect flow-focusing effect, these changes were found to cause variations in jet behavior at each ejection. A circulating refilling system was introduced in place of the conventional tank refilling system. Remarkably, reproducible ejection was achieved for more than 20 minutes. Also, the coating quality are considerably improved than that with the tank system. These results will facilitate the practical use of three-dimensional coating technology of high viscosity-liquids.

    

Microjet Influence on the Mixing of Supersonic Multi-Stream Jet Flow

Active flow control performed on a Multi-Aperture Rectangular Single Expansion Ramp Nozzle (MARS) using a series of microjets perpendicular to the flow. This campaign focused on determining how microjet configuration affects the instability at the interface of the core (M=1.6) and bypass (M=1) streams of the MARS jet. The microjets were located directly at the point of mixing between the two flows as previous simulations determined that point as the most effective at inducing mixing. Three characteristics of the microjets’ configuration were modified: the diameter of the microjets, the spacing of the microjets correlated to a known wave number based on previous passive control simulations, and the angle of the microjets relative to the flows. Actuation of the microjet flow was also be varied. Near field (affixed to the jet’s deck plate) and far field (downstream) pressure measurements taken in an anechoic chamber as well as time-resolved schlieren imaging were used to study the flow. Near field measurements studied the local unsteadiness at the nozzles exit, far field measurements were used to study the general directivity of the acoustic emissions, and time-resolved schlieren imaging was used to visually see how the active control manipulates the overall shock structure.

    

Sliding motion of bubbles in an inclined turbulent channel flow

The sliding motion of a bubble near an inclined channel flow was experimentally investigated to understand the variation of the motion by the inclination in turbulent boundary layers. The inclination of the channel was controlled from 0° to 80°, and Reynolds number of channel flow was 22,000. The qualitative visualization confirmed that there is no significant variation in bubble shape from 0° to 40°, however, bubbles tend to be elongated perpendicular to the wall from 50° resulting from a balance between buoyancy and drag. An optoacoustic measurement technique was adopted, and the optical measurement offers the velocity and diameter of individual bubbles while the ultrasound measurement provides the maximum distance between the wall and bottom of the bubbles with liquid velocity profiles. It was confirmed that the bubble diameter is reduced with the increase of inclination while the bubble height decreases. In addition, the bubble velocity accelerated by the buoyancy, but it slightly increased from 40°. Based on the variables from the optoacoustic measurement, the drag coefficient of the bubble was obtained using a force balance between drag and buoyancy. Finally, we proposed a correlation of drag coefficient using Bond number, Weber number, and ellipticity of bubble.   

Void-wave drag reduction in turbulent flows investigated through different scale experiments

Frictional drag of ships can be reduced by injecting bubbles into the turbulent boundary layer beneath the hull. Author's group has been developing a novel method, repetitive bubble injection (RBI), for improving the efficiency of drag reduction from the conventional method, continuous bubble injection (CBI). The air flow rate of the bubble injection periodically fluctuates in RBI to generate spatiotemporal fluctuation of local void fraction, termed void wave. In this work, the drag reduction caused by the void wave was investigated at two facilities, i.e., a horizontal channel and a towing tank. In the first experiment using the turbulent horizontal channel flow at 5.0-7.0 m/s, RBI at the repetition frequency of 2.0 Hz produces doubled efficiency of CBI at the same bulk void fraction. The second experiment used a 36-m-long flat-bottom model ship towed at 8.0 m/s. RBI at the repetition frequency of 0.5 Hz produces a 13% reduction of the resistance of the model ship and a 3% improvement relative to CBI. The series of experiments demonstrated the superiority of RBI to CBI in terms of time-averaged drag-reduction ratio. We will discuss the factor determining the optimal frequency of void waves for drag reduction based on the experimental results in different-scaled facilities.   

An Experimental Study of Turbulent Noise Reduction through Coupling Owl-like Features

In order to suppress noise associated with high-speed vehicles, owl-wing features that promote silent flight – leading-edge serrations, trailing edge serrations, and the velvet downy feathers – are being explored. The literature indicates very little work done on the coupled effects of these owl-wing features. Thus, the goal of this research is to characterize such experimentally. This is achieved by conducting particle image velocimetry measurements of flow over a plate with serrations at the leading edge, and porous surface conditions (to mimic downy owl wings). The model is tested in a flume at a chord Reynolds number of 65000. The effects are distinguished by comparing with velocities obtained using three other models at similar flow conditions, namely a blunt edge (or the plain) model, a serrated leading-edge model, and a porous surface model. Results indicate the presence of a serrated edge increases momentum thicknesses by over 70%, relative to the plain model. However, by combining the porous surface with the serrated edge, the momentum thickness is enhanced. For the wake flow, vorticity is least in the combined serration-porous surface model, yet with similar turbulence statistics as the plain model. These show that vorticity modification is key to noise reduction.

    

Regular to irregular shock reflection transitions for shock-shock interactions and shock-surface scenarios

Mach stems are generated under certain types of conditions and this can happen when several shock fronts interact, or when a shock front interacts with a surface. A Mach stem generates a localized region characterized by higher pressure, temperature and density when compared to the incident shock front. Understanding when a Mach stem forms and how it propagates is of interest when it comes to, for example, shock-shock interaction events or shock interactions with solid surfaces. Here, Mach stem formation was studied in a laboratory setting using an exploding wire system, which allowed for rapid experiment iteration and exceptional cost effectiveness compared to large-scale experiments or computational modeling. The experimental system is modular and can be easily modified to analyze shock-shock or shock-surface interactions, both of which can be set up as either a 2D or a 3D case. Data was collected via pressure sensors and ultra-high-speed schlieren imaging, and then analyzed to determine the point of Mach stem initiation and Mach stem propagation properties.

    

Returning to the moon safely - cratering and ejecta dynamics during plume-surface interactions

Plume surface interactions (PSI) encompass the process wherein a compressible jet (rocket plume) and a planetary granular surface interacts, forming a crater and dust clouds around landing sites. With NASA’s renewed interest in returning humans to the Moon through the Artemis program, a fundamental understanding of the PSI process will be vital for ensuring the safety of future missions. As such, the primary objective of this work is to study the crater formation process and ejecta dynamics due to PSI by employing non-intrusive optical diagnostic techniques. The experiments are carried out in an atmospheric, bench-scale facility that accommodated a nozzle operating at sonic exit conditions and a sand simulant bed. Stereo photogrammetry is used for quantitative 3D reconstruction of the crater, and planar particle tracking velocimetry is employed to study the ejecta dynamics. Temporal evolution of the crater geometric properties is extracted from the crater reconstruction and compared between different nozzle heights. The ejecta dynamics is also characterized by studying its trajectory angle and velocities. Further analysis of the evolution of these properties at different nozzle heights and the mechanisms contributing to the crater formation process will be presented.

    

Vibrofluidic motion of individual granules

Application of a horizontal, dual mode vibratory waveform has been shown to induce net motion for macroscopic objects, provided that the waveform is non-antiperiodic. Here we present detailed experimental measurements of how the vibratory amplitude, frequency, and phase lag affect the motion of individual granules and larger objects. We show that the net velocity depends strongly on the relative amplitude and phase lag of the second frequency mode. At very large vibratory amplitudes or frequencies, we observe that the velocity is insensitive to the frictional coefficients. We demonstrate that this behavior is consistent with an asymptotic analysis based on a frictional Froude number, and we discuss the implications for control of granular fluids.   

Vibrofluidic pumping, mixing, and separating of granular fluids

We report experimental observations of the behavior of granular fluids when subjected to horizontal, non-antiperiodic vibrations. We present quantitative measurements of pumping velocity, mixing length scale, and separation efficacy as functions of the frictional Froude number, a dimensionless group characterizing the ratio of inertia to the frictional resistance force. We compare the results to computational predictions and asymptotic analyses valid at high Froude numbers, and we discuss the implications for controlling granular fluids in lab-on-a-chip applications.   

Volume reconstruction around bodies for 3D PIV

During 3D PIV experiments, bodies can create fully or partially occluded regions in the measurement volume. By simulating multi-camera PIV experiments with occlusions present, we show that the quality of 3D tracer particle reconstruction depends on the camera configuration and experiment attributes such as the seeding density, particle image radius, thickness of the measurement volume, and the size and brightness of the occlusion. We compare particle reconstruction algorithms to determine which performs best when occlusions are present. We further break down the reconstruction quality by the number of viewing versus occluded cameras and characterize relationships between particle reconstruction quality and velocity field accuracy in partially-occluded regions. Based on these findings, we apply the highest performing imaging and reconstruction parameters from our simulations to an experiment, where they can be combined with experimental improvements, such as fluorescent particles to minimize reflections, to further improve near-body measurements.   

Multiphase Flow Visualization using a Benchtop MRI

Many multiphase flow experimental techniques utilize noninvasive methods like high-speed optical imaging, laser-based diagnostics, and/or X-ray techniques. However, high-speed imaging and laser-based diagnostics cannot see through opaque multiphase flows, and X-rays have challenges with respect to safety and resolution. Magnetic Resonance Imaging (MRI) has been used to visualize multiphase flows, but most MRI machines are often very expensive and/or have limited access. This poster will outline a benchtop MRI that can be constructed with off-the-self items and 3D printed housing. Although much smaller in size compared to a medical MRI machine, it is still capable of multiphase flow visualization. Both gas-liquid (bubble column) and gas-solid (fluidized bed) flows will be visualized using this benchtop MRI.   

Transport of positively buoyant particles in wave run-up on a sloped, porous boundary

While the transport of sediment in waves of nearshore regions has long been of interest, little is known about positively buoyant particles in these systems. For example, coastlines have been observed to act as both a source and sink for microplastics in the ocean, but predicting how these particles are deposited and removed from beaches remains relatively unknown. These systems are highly complicated where the physical processes acting on these positively buoyant particles are influenced by particle size, beach properties, and hydrodynamics. A particular region of interest is the swash zone, where wave uprush and backwash occurs. Uprush and backwash are the upward and downward motions of shoaling waves on a sloped beach. In this poster, we study the transport of positively buoyant particles along a sloped, porous boundary. By generating water bores towards an artificial beach, we can create the upward and downward motion that would transport particles along the slope. In these idealized experiments, we control the size and shape of the particles, the size and speed of the water bore, and the porosity of the artificial beach. Our results illustrate the relative importance of waves, particles, and boundary parameters in this system.   

Data-Driven Rheological Characterization of Thermal Interface Materials

Thermal management plays a crucial role in maintaining optimal temperatures and improving energy utilization of high-power dense electronics and data centers. Electronics thermal management techniques employ thermal interface materials (TIMs) that help reduce junction temperatures and improve heat removal across the solid-solid interface. TIMs are complex paste or gel-like mixtures comprised of a base polymer and conductivity-enhancer particles (ceramic or metallic). However, these complex fluids have complex rheological behavior. Their rheology is non-Newtonian and cannot be represented simply by conventional shear-thinning models. We present a data-driven approach to rheological characterization of TIMs via rheology-informed neural networks (or, RhiNNs, proposed by Mahmoudabadbozchelou and Jalali, Sci. Rep., 2021. The simplest RhiNN considers the TIM as an elasto-visco-plastic (EVP) complex fluid with shear relaxation (via an elastic shear modulus), a yield stress, and a shear viscosity upon flow. Rheological experiments on TIMs under steady and unsteady forcing provide data to feed the RhiNN and solve the inverse problem of rheological parameter identification. Start-up flow experiments are performed at constant shear rates and are repeated for a range of shear rates. Validation of the RhiNNs is carried out by comparing their prediction to experimental data that is not a part of the training set.    

Piezoelectric energy harvesting

Rectangular bluff bodies of uniform area cross-section and attached to two piezoelectric cantilevers are subjected to a uniform flow.  Flow induced vortices cause the piezoelectric cantilevers to vibrate (VIV) in a suction-type wind tunnel that we study to understand influence of different flow and structural parameters. The cantilevered-rectangular bluff body had cross-section length and height ratios B-by-D (B/D) in the range 0.25 < B/D < 0.95. B values between the different B/D ratios are 6.35 mm, 12.7 mm, and 19.05 mm to avoid 5% blockage of the wind tunnel. The thinner (6.35mm) bluff bodies are then susceptible to deformation at higher wind speeds. Between the different combinations of B/D and incident wind speeds, voltage generated ranged from 0.36 V (noise) to 20 V.  Images of the vibrating bluff bodies were captured using high speed video and compared with the voltage levels generated as a function of incident wind speed (1.8-4 m/s). The ratio of natural frequency to vibration frequency proved to be an indicator of operating efficiency.  The operating limit for each bluff body of a given B/D ratio was established in terms of non-dimensional parameters such as Reynolds number (Re)(1000-5300), Strouhal number (St) estimates from the literature, dimensionless mass ratio (1.2-2.6).   

Aerodynamic Loads on Simultaneously Pitching and Rotating Flat Plate: Effect of Pitch Rate and Pivot Location

The flow over a rotating and pitching wing is unsteady, three-dimensional, and a simplification of insect wing motion. The design of bio-mimicking micro air vehicles is inspired by this motion, which sparks interest in studying flow structures along a simultaneously pitching and rotating flat plate. The influence of wing pitch rate and pitch pivot location on aerodynamic performance of rotating wings has not yet been fully investigated. To this end, the present investigation utilizes a flat plate (aspect ratio = 6), rotating at a Re of 20000, and pitch angle ranging from 0 – 90°. Two parameters are considered in this poster: the non-dimensional pitch axis (x_p/c = 0, 0.25, 0.50, 0.75, 1) and the reduced pitch rate (0.03 ≤ K ≤ 0.3); which is the ratio of pitch velocity to translational velocity (K = α^'c/ 2U_tip). Qualitative flow visualization was conducted via fluorescent dye at the radius of gyration of flat plate. Wing loads were measured using a multi-axis strain gauge load cell. The preliminary results show a generalization in scaling of lift and drag with reduced pitch rate and pivot point and relate such scaling to the formation, growth, and pinch-off of leading-edge vortices.

    

Two-body freefall time in a matter dominant expanding universe

Predicting freefall time is a major problem in galactic dynamics. The relaxation time of collisionless dark matter, which heavily influences galaxy evolution, is closely related to its freefall time. Calculating freefall time using Kepler’s Third Law does not consider the effects of the universe’s expansion. This paper proposes a new formula for calculating freefall time in an expanding background. The derivation of the formula starts from the governing equations of an N-body system in an expanding background. Two-body simulations were run using equations of motion with constant damping (to represent the effects of expansion) derived from the N-body equations. More than 5,200 different two-body systems with different separation distances, mass, and damping were simulated. The simulated freefall time was compared with good agreement against the proposed formula. The results demonstrate that two-body freefall time is dependent on the time at which collapse began. The earlier it began, the longer the freefall time. A weak gravitational attraction, caused by the small mass of the system or the great distances involved, exacerbates this effect. The galaxy that started to form earlier is expected to take longer to form. A better understanding of freefall time can lead to greater insight into galaxies’ historical and future evolution.    

Machine Learning-Based Three-Dimensional Reconstruction of Stress Field from Flow Birefringence

In biological fluid dynamics, measurement of the stress field in a blood flow is crucial for various applications, such as understanding the cause of cerebral aneurysms. However, since blood is a complex fluid, the constitutive equation for computing the stress field from the velocity field is still unclear. Therefore, we proposed a machine learning-based photoelasticity approach to measure the stress field of a complex fluid flow. Here, we show the results of the three-dimensional reconstruction of stress field from the measured flow birefringence. Numerical simulation was performed to obtain the data of the flow birefringence and stress fields of an axisymmetric pipe flow. The data is then used to train a deep convolutional encoder-decoder (DCED) model to predict the stress field from the flow birefringence. DCED achieved high accuracy in predicting the simulation data. Even for the experiment data, it has successfully predicted the trend of the stress distribution across the pipe's cross-section. The prediction can be improved by including the experiment data into the training data. The results obtained in this study indicate a major advance towards the non-contact stress measurement of a fluid flow, which will be an important tool to the field of biological fluid dynamics.   

A multiphase approach to model bacterial biofilm

In this work, we use a modelling and experimental approach to study the self-organization of bacterial cells expanding on solid agar. Growing microbes form a biofilm by secreting extracellular polymeric substances (EPSs) that form an integral part of biofilm-like structures. In these structures, interaction among mechanical and physical forces results in different morphological patterns such as concentric rings, dendrites, Eden like structures. Surfactant molecules that lower the surface tension of the colony are secreted by these growing cells. Lowering the surface tension helps spreading of bacteria on solid surfaces. The difference in the surface concentration of surfactant molecules causes the Marangoni effect that facilitates spreading.

In this work, we study the effect of bacterial surfactant secretion, the diffusional flux of nutrients, substrate thickness and wettability on colony growth and shape within the framework of a hydrodynamic thin film model. Our framework accounts interplay among mechanical stress from bacterial migration, growth, surface stress and Marangoni effects. Our mathematical model is able to explain front instability in bacterial colony patterns and various experimental observations of microbial colony patterns in literature.

    

Most forward flying insects do not cruise at "optimal" Strouhal numbers

Insects are the smallest flying animals, representing nature's solutions for the problem of flapping flight at Reynolds (Re) numbers in the range of 10   

Flying spiders: Effects of the spider mass and the dragline length in spider landing

Many spiders use aerial dispersal "ballooning" to move from one location to another. By ballooning, spiders can reach distances as far as 3200 km and heights up to 5 km. Although this process is exhibited regularly in nature, little is known about the dynamics at play. What dominates the three stages of spider takeoff, flight, and settling? Understanding the roles of multiple parameters, including a spider's mass, morphology, posture, dragline properties, and local meteorological conditions (turbulent level and thermal stability, is not only of ecological significance but critical to improving advanced technologies for bio-inspired innovations of airborne robotic devices. Here we pose a controlled lab experiment to test for the characteristics of free fall landing. We seek to determine how the dragline length and spider mass affect the dynamics of freefall at Reynolds numbers of several thousand using high-speed recordings. Such results are expected to shed light on the intriguing flow physics of spider ballooning and validate new models.

    

Microfluidic cell-tracking technique in a comoving frame during pulsatile capillary flow

One of the most intriguing features of red blood cells (RBCs) is their high deformability, which allows them to squeeze through microvessels smaller than the RBC size. Additionally, this flexibility results in a broad range of RBC shapes in microfluidic flows, depending on their confinement and flow velocity. RBCs are often recorded at a fixed channel position using optical microscopy to study their flow and shapes in microfluidic devices. However, this approach limits investigations of the RBC flow behavior and shape transitions to merely a few hundred micrometers, depending on the used camera and objective.

This study presents a microfluidic technique that allows us to track RBCs over several centimeters along the channel flow direction. The microfluidic setup consists of an inverted microscope equipped with a motorized stage, a 20x objective, a highspeed camera, and a high-precision pressure device to generate a time-dependent driving of the flow. We use a customized feedback-control mechanism between the stage and camera to obtain the comoving frame. RBCs are instantaneously tracked, and the observation frame follows the cells in the microfluidic device with a velocity of up to 10 mm/s. Our method enables us to probe the flow behavior and shape transitions of individual RBCs along the channel flow direction under steady and pulsatile flow. Thus, the presented technique opens new pathways to study cell dynamics and interactions in microfluidic devices under physiologically relevant flow conditions.

    

Vortex patterns in the wake of pitching, heaving, and combined motion bio-inspired pitching panels

An understanding of how bio-inspired motion interacts with a surrounding fluid environment provides a foundation to supplement a lack of sensory data and learned intuitive motion in synthetic underwater vehicles. Prior work within our research group has focused primarily on the wake dynamics and propulsive performance of isolated, purely pitching propulsors with bio-inspired planforms. The current work builds upon prior results by experimentally investigating pure heaving motions and combined pitching and heaving motions for a series of bio-inspired fin planform geometries. Stereoscopic particle image velocimetry collected in the midspan at a Strouhal number of 0.3 is used to explore the time-varying behaviors of wake vortices. The current work investigates the influence of both propulsor planform and kinematics on wake dynamics, and how that corresponds to what is known in the literature about the timing and magnitude of the circulatory and added-mass contributions to the thrust force and propulsive efficiency.   

Effects of prey capture on the swimming and feeding performance of choanoflagellates.

Choanoflagellates, eukaryotes that are important predators on bacteria in aquatic ecosystems, are used as a model system to study the evolution of animals from protozoan ancestors. The choanoflagellate, Salpingoeca rosetta, has a complex life cycle that includes unicellular and multicellular stages, provides a model system to study the consequences of different cell morphologies. A unicellular S. rosetta has an ovoid cell body and a single flagellum surrounded by a collar of microvilli. The cell swims by waving its flagellum, creating a water current that brings bacteria to the collar of prey-capturing microvilli. One measure of the performance of a suspension-feeding organism is the volume of fluid that it can move into its collar during a beat cycle. The inward flux of fluid acts as a proxy for the rate of bacterial capture. Here we use a regularized Stokeslet framework to model the hydrodynamics of a unicellular choanoflagellate, the captured bacterial prey, and their effect on swimming performance and clearance rate. We compare model predictions with high-speed microvideography. Moreover, we will discuss current assumptions, and future model improvements that, with coordinated lab experiments, will help us probe this intriguing biophysical system.   

Building a macroscopic three-link swimmer at low Reynolds number

Purcell theorized that one of the simplest robots that can move at low Reynolds number is a three-link swimmer, which consists of three hinged links in a chain. The work of Hatton and Choset (IEEE Trans. Robot, 2013) provides a theoretical framework to predict the displacement and rotation of such a swimmer, but assumes the robot links are slender. They use a phase space of the two joint angles and a 3-D map known as a height function derived from the Navier-Stokes equations to determine the motion of the swimmer. However, no one has tested this theory experimentally using a macroscopic three-link robot. Our robot will swim in highly viscous silicone oil so the Reynolds number is small, and we track the motion in videos to compare the displacement and rotation to that predicted by theory. We will also determine the height function for our robot empirically (Hatton et al., PRL 2013) to understand the error introduced by our robot not being a slender body. We use a 3-D printed body and vary its shape to see how the how the performance of the swimmer changes.   

Calibrating numerical simulations with macroscopic bacterial models

The swimming of microorganisms is typically studied using biological experiments and/or numerical simulations. However, numerical simulations of microorganisms are often not compared to precise measurements because of the difficulty of making microscopic measurements of forces and torques in biological experiments, which are typically ∼ 10 µm. Instead, our research group uses robotic models that are about 10 cm in size and match the Reynolds number of swimming microorganisms by using highly viscous silicone oil that is 100,000X more viscous than water. We can then measure the translational motion of the models and scale the results from our dynamically similar experiments to biologically relevant sizes. We have used other experiments to calibrate the method images for regularized Stokeslets and found excellent agreement between our data for both cylinders and helices. Previous results also confirmed the theory of Jeffrey and Onishi (1981) for the torque on a cylinder near a plane wall, as reported in Shindell et al., Fluids (2021). Our current work tests the theory of O’Neill (1964) for a sphere moving near a wall as another reference for establishing appropriate parameters for Stokes flow simulations.

    

Cardiovascular Flow Diagnosis using Integrative 4D MRI Scans and Virtual Reality System

Cardiac Magnetic Resonance Imaging (MRI) uses magnets and pulsed radio frequency to create detailed computer images used to diagnose and screen patients with cardiovascular diseases. The 4D flow MRI is a recent cutting-edge imaging technology that offers more detailed information about the heart chambers and the arterial network. This technology allows clinicians to better visualize blood flow through the cardiovascular system. Although the 4D flow MRI provides clinicians with detailed blood flow structures, it may be more optimal when coupled with real-time immersive visualization technology, like Virtual Reality (VR) systems. In this work, we developed an integrative platform that combines 4D flow MRI data sets with a state-of-the-art VR wireless system that allows clinicians to interact with these data sets using a headset and controllers. The platform is not only flexible enough for visualization purposes, but it can also be used to extract quantitative features for accurate cardiovascular diagnostics.   

Estimation of myocardial material parameters in developing zebrafish using inverse finite element analysis

The zebrafish is a widely used model for studying cardiac development. It is well known that mechanical forces are essential stimuli for the proper development of the heart. Multiphysics in silico models of the zebrafish heart, combining fluid-structure interaction, electrophysiology, and micromechanical tissue models, are powerful tools to quantify mechanical forces with high spatiotemporal precision, and they promise to greatly increase our understanding of cardiac mechanobiology. However, such models rely on accurate estimates of the material parameters of the zebrafish heart wall, which are difficult to obtain experimentally due to its small size. We address this limitation by applying an inverse finite element (FE) analysis to estimate the in vivo material parameters of the zebrafish ventricle. First, an image-derived FE model of the ventricle is coupled to a 0-D model of blood flow to simulate diastolic filling. Then, material parameters of the FE model are found using an optimization routine by matching simulation results to in vivo videos of the zebrafish ventricle. With the optimized parameters, the model shows good agreement with the in vivo deformation of the ventricle, demonstrating the viability of our approach. This work is supported by grants from the NIH and NSF.   

Simulated Performance of a Bioprinted Pulsatile Fontan Conduit

Fontan physiology in single ventricle patients causes circulatory inefficiency, which contributes to various morbidities. To resolve these unfavorable hemodynamics, our interdisciplinary team is developing a 3D bioprinted conduit that could theoretically contract to provide a pulsatile energy source to the lungs. In this study, we evaluated design parameters of the pulsatile conduit by creating a 3D finite element simulation framework. We combined electromechanics with fluid-solid interaction in the 3D model and coupled it to a close-loop lumped-parameter network (LPN) of the Fontan circulation. This was implemented in the open-source software SimVascular. We used this framework to evaluate the effect of varying the conduit geometry, Purkinje network and fiber direction. The conduit was assumed to generate a 5 kPa active stress with 3 kPa passive stiffness based on previously estimated values of tissue-engineered materials. In a conduit with purely circumferential fiber alignment, we observed a 13. 61% volume amplitude while a 4 mmHg increment in conduit pressure. The central venous pressure of the Fontan circulation decreased from 15.5 to 14 mmHg. In a conduit with asymmetrically-oriented fibers, we observed a twisting motion as the dominant motion in the ventricle. This model generated a 9.3% ejection fraction and 2.5 mmHg pressure elevated. The central venous pressure could be able to decrease 1.4 mmHg. We concluded that the current performance of conduit is not sufficient to significantly improve Fontan physiology performance and recommend alternative target parameters that would yield desired outcomes.   

Characterization of the developing venous vascular system and its connection to cardiac outflow hemodynamics

Congenital heart defects involving venous malformation is a complex class of diseases with largely unclear etiologies. Here, through hemodynamic simulations, we study the development of the cardinal and pharyngeal veins and how they are altered by perturbations to systemic circulation by means of pharyngeal arch arteries (PAA) ablation. Using the chick embryo animal model, we obtain 3D reconstruction of PAAs and surrounding veins from high-resolution nano-CT scans. We perform multiscale computational fluid dynamic simulations of day 3 and day 4 embryos using the open source platform SimVascular. We use Doppler velocity and pressure measurements to generate the appropriate inlet and outlet boundary conditions. From here we calculate flow distributions and wall shear stress maps. We perform the analyses first on control embryos to establish the natural variabilities in venous circulation. Then, we repeat the analyses on embryos with ablated PAAs to characterize venous response to such mechanical alteration of systemic flow. This study addresses the interdependence of the developing systemic and venous vascular systems and serves as a basis for understanding the role of blood flow in venous growth, adaptation, and malformation.   

Towards a rapid extraction of three-dimensional Lagrangian coherent structures from in vivo 4D-MRI

Interventricular fluid dynamics are complex and three-dimensional. Moreover, under pathological conditions, such as aortic regurgitation, the complexity of the ventricular flow requires fast and efficient three-dimensional analysis tools. Therefore, this study aims to extend the application of Lagrangian descriptors to identify three-dimensional coherent structures in cardiovascular flows. Using in vivo 4D-MRI velocity fields, ventricular coherent structures in the presence of aortic regurgitation are compared to structures in healthy patients. The results show that the method based on Lagrangian descriptors can rapidly extract flow coherent structures when compared to other approaches. Moreover, the identified structures in the presence of aortic regurgitation reveal the breakup of the optimized left ventricular filling structures.

    

Network-based study of Lagrangian trajectories to highlight the effect of bioprosthetic mitral valve oreintation

During the ventricular diastolic phase, the blood flowing through the mitral valve preserves its kinetic energy by forming a large vortex filling almost the whole left ventricle. This vortical structure, then, redirects the blood towards the aorta for the ejection phase. In the presence of mitral valve disease, valve replacement with a bioprosthetic valve becomes an option for symptomatic patients. Surely, the orientation of the implanted bioprosthetic valve with respect to the ventricular apex will affect the energy-preserving behavior of blood transport. Therefore, this in vitro study investigates the effect of three different orientations of valve implantations on ventricular fluid transport. Planar time-resolved particle image velocimetry is used to measure the instantaneous velocity fields at several heart rates. Then, using a network-based analysis of Lagrangian trajectories we reveal the interventricular blood transport mechanism and how it is affected by the orientation of the valve. Also, by using a spectral analysis of the network, we reveal the major transport clusters in the flow field. Finally, our study highlights the overlap between the identified clusters and the Lagrangian coherent structures in the flow as extracted using finite-time Lyapunov exponents.   

Margination of discoidal micro-sized particles in blood flow

Micro-sized vascular-targeted drug carriers and imaging agents are promising for therapeutic applications. The hydrodynamics-driven migration of these micro-particles toward the vessel wall is known as margination and is the prerequisite of adhesion to the endothelial layer. Herein, we apply the lattice Boltzmann method coupled with the immersed boundary method to investigate the effect of the particle properties, vessel size, and hematocrit on the margination of discoidal polymeric micro-particles in whole blood. Experiments have shown that Leukocytes account for Young's modulus dependence of the adhesion of micro-sized particles on the vessel wall under different shear rates. On the other hand, micro-sized drug carriers can alter the dynamics of white blood cells in blood flow. We exploit our cellular-sale modeling to examine these complicated dependences. Besides, red blood cells rigidify and undergo morphological changes in diseases. We also look into the effect of red cell deformability on the margination of white blood cells and micro-sized drug carriers.

    

Regional variation in homeostatic stretch and stiffness in a swine model of branch pulmonary artery stenosis

Branch pulmonary artery stenosis (PAS) often presents in children with congenital heart defects, altering blood flow and pressure in the pulmonary circulation. Perturbations from homeostatic arterial mechanics have been proposed as stimuli for vascular growth and remodeling. However, there is a current lack of quantification of these stimuli in the pulmonary arteries and their alteration during pathology. In this study, we used a 1D fluid dynamics model to simulate pressure-flow relationships in the pulmonary artery tree. Left PAS was surgically created in 4 swine at 2 weeks of age, with 4 additional sham animals. Pressure and imaging measurements were obtained at 20 weeks of age. The unloaded radius of each vessel was determined from the model predicted pressure-area relationship. In each sham animal, our results indicate little variability in regional homeostatic stretch when we prescribed constant arterial stiffness. In addition, there was little inter-animal variation in fitted stiffness or stretch. In contrast, we expect regional differences in stretch and stiffness will be necessary to match the PAS experimental data and that differences from sham will indicate arterial growth and remodeling.

    

A 1-D Pulse Wave Model for Perivascular Flow

In vivo measurements of cerebrospinal fluid in the perivascular spaces (PVSs) indicate that flow is correlated with arterial wall motion. However, 3D simulations of the network are computationally expensive. To address this, we apply a one-dimensional modeling approach that couples the perivascular and arterial networks through vessel wall deformations that allow for wave reflections and transmission at vessel boundaries. We assume area waves from the arterial network are coupled to the PVS network’s area in a one-way interaction, a three-element Windkessel model represents the microvasculature, and pressure and volumetric flow are continuous at bifurcations. Our solver uses Richtmyer’s method, which is second-order accurate in time and space. This approach allows arterial deformations to be explicitly calculated for each segment in the entire network, rather than relying on isolated in vivo measurements of wall motion to drive perivascular flow.

    

Influence of viseme-inspired orifice shapes on the dispersion of expiratory events

It is well established that diseases can be spread via airborne transmission, sourced from expiratory events such as speech, coughing and sneezing. As these events are pulsatile in nature, it is important to study the influence of pulsatility on the flow dynamics of these events. In this work, the influence of mouth shape (inspired by visemes, i.e. sounds that look the same) on the dispersion of pulsatile expiratory aerosols is studied to help understand and mitigate the spread of airborne diseases. A custom-built pulsatile coughing simulator is used to generate single- and multi- pulsed coughs of varying cough peak flow rates (CPFR). Five static, interchangeable elliptical orifice shape fittings (e = 0, 0.70, 0.85, 0.95, 0.99) representing different visemes are 3D printed and attached to the outlet of the simulator. Each fitting has a cross-sectional area equivalent to a circle with diameter D=1in. The penetration distance of both single- and double-pulsed coughs are measured using smoke flow visualization for each case. Particle image velocimetry is used to measure the exit velocity profile and maximum axial velocity over time at the mouth outlet. It is observed that the axial exit velocity and penetration distance are both higher for orifice shapes with larger eccentricities.   

Computational modelling of respiratory lung networks to predict patient-specific ventilatory responses

Mechanical ventilation system is a lifesaving treatment for patients who develop severe respiratory problems including acute lung failure (Acute Respiratory Distress Syndrome, ARDS). Although the ventilation system can be employed by the medical teams to keep a paitient's lung open to ensure the continuous exchange of oxygen and CO2, at the same time the ventilation pressure can cause severe damage to the lungs that it results in the patient's death. Doctors treating patients for acute respiratory problems have a limited range of parameters to work with when determining the best protocol for mechanical ventilation – pressure limits, oxygen level and air flow, for example. But the lung is a complex organ, and the amount of pressure necessary to keep all parts of the lung open to airflow can actually cause damage to some parts through overdistention of the tissue. In this study, we aim to develop a computational model to simulate the tempo-spatial variation of pressure and volume that occur during an inhalation and exhalation cycles. The computational model proposed in this study is formed around multiple algorithms developed using the concepts of pore network models for multiphase systems. The digital lung models are generated based on CT lung scanning while the simulation results can predict the response patient-specific ventilatory responses including the pressure fields along the entire lung system. The outcome of this work, can help medical teams to treat patients for acute respiratory problems by optimizing the mechanical ventilation systems.   

Topological transitions in fluid lipid vesicles: activation energy and force fields

Topological transitions of fluid lipid membranes are fundamental processes for cell life. For example, they are required for endo- and exocytosis or to enable neurotransmitters to cross the neural synapses. Inspired by the idea that fusion and fission proteins could have evolved in Nature in order to carry out a minimal work expenditure, we evaluate the minimal free energy pathway for the transition between two spherical large unilamellar vesicles and a dumbbell-shaped one. To address the problem, we propose and successfully use a Ginzburg-Landau type free energy, which allows us to uniquely describe without interruption the whole, full-scale topological change. We also compute the force fields needed to overcome the involved energy barriers. The obtained forces are in excellent agreement, in terms of intensity, scale, and spatial localization with experimental data on typical fission protein systems, whereas they suggest the presence of additional features in fusion proteins.   

DNS Study of Drag Reduction Effect on Ultra-Fine Rough Surfaces

The ultra-fine rough surfaces, even if it is randomly-distributed in space, sometimes delay transition and reduce drag, when their size is within the viscous sublayer (Tani (1988), Oguri et al. (1996), Tameike et al. (2021) and Hamada et al. (2022)). We name that surface structure to reduce drag as Distributed Micro Roughness (DMR). We demonstrated direct numerical simulation of Tollmien–Schlichting (T-S) transitional flow over the sand-grind roughness surface, that was implemented in the real shape we used in the wind tunnel experiment. Firstly, we confirmed the drag reduction effect over the sand-grid rough surface, by observation of the turbulent kinetic energy (TKE) distribution and friction drag with calculation of the integral value of the averaged Reynolds shear stress. Secondary, we analyzed the mechanism of the statistic change, by applying the statistical decomposition of temporal phase-averaging of the TS wave and other three-dimensional components. The results show that the TS vortex tends to break down into turbulence over the sand-grind rough surface, although TKE and friction drag were suppressed compared to the smooth surface. In the presentation, we will discuss the mechanism more in detail.   

Pulsating Turbulent Channel Flows over Sandgrain Roughness

In this work we investigate the behavior of oscillating channel turbulence over a rough surface by running a series of DNS under $Re_\tau = 1500$. The virtual sandgrain roughness used consists of closely-packed ellipsoids with random orientations. The roughness height is 4\% of the channel half height. Pulsation of the flow is applied by a sinusoidal variation of the pressure gradient around the mean value giving the target $Re_\tau$. Three frequencies in the very high, high, and intermediate regimes are examined. The time-averaged centroid of the force exerted by the roughness is the same as it is in the non-oscillating case. The equivalent roughness height remains the same for the intermediate- and low-frequency pulsation while increasing significantly for the high-frequency one. Compared with the canonical Stokes layer whose thickness is $2l_s$, the oscillation wave propagates further into the channel over the rough wall to $10l_s$. This range does not vary with the change of the oscillation frequency for the roughness-forcing combinations considered here, and it is still within the roughness sublayer ($y<3k$). The local wake flow around the roughness elements responds actively to the pulsation and leads to a significant variation in the wall stress during each cycle.

    

Dual-Tracer Laser-Induced Fluorescence Thermometry for Understanding Bubble Growth during Nucleate Boiling

Boiling heat transfer associated with bubble growth is perhaps one of the most efficient cooling methodologies due to its large latent heat during phase change. Despite significant enhancement of heat removal rates, numerous questions remain regarding the fundamentals of bubble growth mechanisms, a major source of enhanced heat dissipation. This work aims to accurately measure three-dimensional (3D), space- and time-resolved, local liquid temperature distributions surrounding a growing bubble that help better explain the heat transfer to bubble growth. An artificial cavity of  in diameter is fabricated on a rectangular-shape heat sink as nucleate sites. The dual tracer laser-induced fluorescence (LIF) thermometry technique is combined with a high-speed imaging method to capture transient temperature distributions of the single bubble. This technique successfully provides fluid temperatures with unprecedented accuracy at micrometer resolution. It measures two-dimensional (2D) bulk fluid temperature fields within 0.3 ºC at a 30 μm resolution. Two temperature-sensitive fluorescent dyes, Fluorescein (FL) and Sulforhodamine B (SrB) are used as temperature indicators in the LIF technique to have an accurate measurement. A laser light sheet scanned across the entire measurement volume excites the fluorescent dye, and an optical system involving a color beam splitter gives the intensity distribution of the individual fluorescent dyes on a high-speed camera. The temperature data is used to quantify time-resolved heat fluxes contributing to mass transfer near the growing bubble. 

    

Thin film evaporation modeling of the microlayer region in a dewetting water bubble

Understanding the mechanism of bubble nucleation and growth is critical for optimizing the boiling heat transfer process. The microlayer region formed during bubble departure is known to be a major contributor to the phase change heat transfer but its evolution, spatio-temporal stability, and impact on macroscale bubble dynamics are still relatively unknown. In the thin film microlayer region, the vapor pressure is balanced by a combination of the capillary, disjoining, and liquid pressure. Thermally driven evaporation in the curved, thin film is suppressed by disjoining pressure at the nano-scale leading to a peak in evaporation flux in the microlayer region. Using the conservation equations with a lubrication approximation coupled with the augmented Young-Laplace equation, 3rd order nonlinear film evolution equation is obtained. The evolution equation solved numerically using the ODE 45 method in MATLAB. A variable wall temperature BC is employed at the solid-liquid interface and is balanced by evaporative heat loss at the liquid-vapor interface. Contrary to most models, the solution begins in the thicker region of the film and proceeds in a direction of reducing film thickness. The adsorbed film thickness is not known a-priori but is an output of the model. The solution method is devoid of tuning parameters or guessed inputs. Experimentally measured film thickness and its derivatives are used as inputsin the thicker region and the model is evaluated to obtain a film profile down to the nano-scale and a corresponding local evaporative flux. The modeling results compare favorably with spatio-temporal experimental measurements of the microlayer. By comparing the film profiles at multiple time steps, the accommodation coefficient could be determined and its influence on the evaporative flux distribution is discussed.   

Implicit large-eddy simulation of subsonic compressible low Reynolds number flow over a flat plate at Re = 20,000

We performed an implicit LES around a 5% thick plate with a Reynolds number of 2.0 × 10⁴ and investigated the freestream Mach number effect on the flow structure and turbulent transition. As in the Reynolds number of 1.1 × 10⁴ reported in the previous studies, the turbulent transition is delayed, and the length of the laminar separation bubble increases due to an increase in the freestream Mach number. The tendency was remarkable at the Mach number of 0.5. In the case of Mach number of 0.8, the pressure plateau disappears, and flow does not transition to turbulence at the Reynolds number of 1.1 × 10⁴, but it was found that the reattachment occurs after the turbulent flow transition when the Reynolds number is 2.0 × 10⁴. In addition, the transition point and the reattachment point move downstream as the freestream Mach number increases, and the distance from the transition point to the reattachment point is also increased by increasing the freestream Mach number. This result indicates that the turbulent transition gently occurs under the high Mach number condition. In addition, the peak Reynolds stress is reduced under high Mach number conditions.

    

Simulation of Cavitation during Shock-Droplet Interaction

This study aims to model cavitation that occurs during a shock-droplet interaction using industry-standard cavitation models such as the Schnerr-Sauer model and Rayleigh-Plesset equation available in STAR-CCM+. The underlying motivation for this study is to investigate how cavitation will affect droplet breakup when interacting with strong shock waves from vehicles during reentry. Three Eulerian phases are simulated with air and water vapor being modeled as ideal gases while liquid water is modeled using second-order compressible equations of state. Simulation data is then validated against previous experiments.

    

Modeling of sub-grid scale eddies using machine learning

There are two main computer models for investigating turbulent flows, the most accurate of which is called Dynamic Numerical Simulation (DNS). This model resolves all scales of eddies directly, however, it requires the large number of grids and spends much time. The other model as an alternative to the former one is Large Eddy Simulation (LES) which resolve the larger-scale eddies directly and requires artificial modeling for the remaining smaller-scale eddies. Therefore, the results of the LES model can vary depending on the sub-grid model for modeling smaller-scale eddies. To make the performance of the LES model as similar as possible to that of the DNS, the correlation between large and small eddies is found using deep learning.

This study uses two representative deep learning models, Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN), which is based on the ANN algorithm. RNN has been used as a standard method for machine learning of sequence data (audio, natural language, etc.) (Yin et al., 2019). So, RNN needs a good memory which remembers all input in internal memory to forecast the future events by reminding previous data. While, CNN is a powerful tool to extract fluid dynamics features and predict flow fields. Strictly speaking, CNN recognizes geometrical topologies of jet flow. In this study, the flow structure characteristics are investigated using CNN, and the flow over time is examined using RNN.   

Computational assessment of distribution of plate-type windborne debris

In the severe wind such as typhoon, major wind damages are caused by windborne debris, some of which results from the damaged structures, others from loose-laid objects. Due to limitations of previous experiments and resulting analysis, relatively uniform wind has been considered, whereas three-dimensionality of surrounding wind field is very important to trace the trajectory of windborne debris. In this paper, computational investigation is presented to demonstrate the importance and the role of the three-dimensional wind field in the trajectory of windborne debris. Unsteady Navier-tokes equation is solved for the turbulent flow around a three-dimensional structure. Loose-laid plate-type windborne debris is launched at a prescribed wind speed and the trajectory is computed based on the equation of motion and 3D turbulent wind velocity computed from the simulation. Interests are put on the effects of boundary layer on the trajectory and the comparisons are made between the trajectories in the uniform flow and the boundary layer. The results show that the trajectory is heavily dependent on the vertical component of wind velocity and the vertical gradient of inflow.   

A computational anaysis of heat enviroment around unban pavement

Due to climate change and urbanization, occurrences of heat island and tropical night are increasing as concrete structures and manmade roads have been extended and green land has decreased, resulting in loss of even human lives and damages on facilities. Application of special material to pavement has been researched to utilize its effects of lowering temperature and high capacity of water retention. In this study, computational study were undertaken to analyze the heat balance on and in the pavement of materials water retention property. The computations were based on computational fluid dynamics with different values of the water content percentage and physical properties. The result shows that the physical properties of pavement doesn't affect the surface temperature of the pavement, while increasing water content percentage in the material upto 50% has lowered the surface temperature by 3? and spraying water on the pavement has lowered the temperature by 11?. However, in spite of spraying water, the air temperature near the pavement remained higher than 25?.   

Towards exascale multiphase compressible flow simulation via scalable interface capturing-based solvers and GPU acceleration

OLCF Frontier broke the exascale barrier in June 2022.

One must judiciously offload and carefully implement CFD algorithms to effectively use exascale resources for flow simulations. 

We present a strategy that brings us closer to utilizing these resources in the context of multiphase compressible flows.

All implementation is in the Multicomponent Flow Code (MFC; open source).

MFC uses a finite volume method with WENO5-based interface capturing and the HLLC Riemann solver.

We offload all compute kernels to GPUs via OpenACC and argue that similar explicit methods on current accelerators must tack this way.

The kernels are fine-tuned via OpenACC decorations and thread serialization where appropriate.

This results in high arithmetic intensity, realizing about 55\% of the peak GPU FLOPs.

We observe a 500-times speed-up on an NVIDIA A100 over a single core of a modern Intel CPU.

This corresponds to about a 50-times speed-up over a CPU node for a  GPU node in a modern supercomputer (e.g., SDSC Expanse).

This implementation demonstrates ideal weak scaling to at least 13824 GPUs on OLCF Summit.

CUDA-aware MPI enables remote direct data access, improving strong scaling behavior.

We probe performance against various CPU architectures, including ARM, x86, and Power   

Characteristics of Wakes from 3D Rotating Ellipsoids using Nonconforming Schwarz-SEM

Vortex shedding and the related added-mass effect’s impact on drag and lift felt by rotating spherical and ellipsoidal bodies moving in flows at moderate Reynold’s numbers has been recently quantified using a novel moving nonconforming Schwarz-spectral element method. In the current study we focus on the characteristics of the vortices generated by rotating ellipsoids of different aspect-ratios, using direct numerical simulations (DNS). We capture the toroidal vortices generated, whose shape changes based on the aspect ratio of the ellipsoid. Significant difference in the wake is observed between the rotating and non-rotating case.

    

Experimental Investigations on the Hydrodynamics of an Idealized Wildland-Urban Interface (WUI) with Implications to Wildfire Management

The wildland-urban interface (WUI) is characterized by a sharp transition between vegetated land and non-vegetated or paved land. The change in land conditions affects the propagation of wildfires across the interface, which is known as the edge effect in the literature of vegetated canopy flows. Vegetation plays a critical role in determining the propagation speed and the direction of fire. It is well-known that fire spread (speed) positively correlates with wind speed (Linn et al. 2002 and 2005). The latter, in turn, is governed by the interactions between the approaching wind, the canopy structure, and the fire (Keeley and Syphard (2019)). Thus, any predictions of fire spread must begin with the understanding on the fluid dynamics of canopy flows. A conundrum in fire containment is that the clearing of trees to create an area void of fuel (fuel-break) is not always effective in stopping the advancement of fire (Pimont et al. (2009)). In some cases, it actually accelerates fire propagation. This conundrum can be resolved by realizing that the supply of heat is governed by canopy flows and that the clearing of trees speeds up the local wind such that heat from one location can reach another location with trees (fuel). We present a preliminary experimental study to address two hypotheses: (1) near the ground the spread of fire is mostly governed by horizontal mixing layer whereas at the top of canopy both vertical and horizontal mixing layers are active, and (2) the multiscale nature of trees enhances turbulent mixing and results in a faster fire spread. Our experiments consist of a lab-scale fractal tree canopy model placed inside a water flume and point buoyant sources are distributed within the canopy to mimic fire behavior. We considered two canopy parameters – canopy density and the presence/absence of a midstory . Planar particle image velocimetry (PIV) and laser-induced fluorescence (LIF) were used to quantify turbulent mixing of a passive tracer (heat).   

Mixed convection in an idealized coastal urban area with multi-scale surface heterogeneity.

Marine heatwaves (MHWs) modulate the coastal climate through interacting with local circulations, such as meso-scale land-sea-breeze circulation and micro-scale urban-rural circulation. These circulations result from the meso-scale surface heterogeneity between land and sea surfaces, and micro-scale surface heterogeneity between the urban built-up areas and the surrounding rural areas, respectively. In this study, we consider the turbulent mixed convection within the planetary boundary layer (PBL) of the coastal region, and investigate how two different scales of surface heterogeneities impact the mix convection. The MHW is modeled as a turbulent inflow with elevated air temperature. Informed by dimensional analysis, four non-dimensional parameters are considered for a fixed setup of the urban-rural-sea system, in which two of them are related to the relative inflow speed and temperature , and the other two are related to the relative buoyancy flux of urban, rural and sea areas. Then, a suit of numerical experiments are carried out using large-eddy simulations (LESs) to systematically study the impact of the above four non-dimensional parameters. We find that the MHW, land-sea-breeze circulation and urban-rural circulation interact synergistically. The temperature difference between urban air temperature and inflow temperature is larger in the case with elevated inflow air temperature comparing to that of the base case with low incoming air temperature. It also increases with decreasing inflow velocity. The elevated urban air temperature will further strengthen the land-sea-breeze circulation. This study highlights the importance of multi-scale surface heterogeneity in modulating mixed convection.

    

Effect of Rayleigh and Prandtl Numbers upon Natural Convection and Flow Field in a Differentially Heated Cavity

In this poster presentation, direct simulation of 2D and 3D differentially heated square cavities (DHC) filled with fluids of Prandtl (Pr) numbers ranging from 0.5 to 100 is conducted to explore and visualize the influence of the Prandtl and Rayleigh (Ra) numbers upon the temperature and velocity fields within the cavity. 

It is observed that for constant Ra, the velocity and temperature fields between low Pr and high Pr fluids differ significantly, especially near the corners of the cavity due to inertial effects. These inertial effects are found to drastically impact the temperature and velocity fields for Ra larger than 10⁵. For Ra > 10⁵, there is a clear divergence between the flow fields of Pr  < 1 and Pr > 1 as increases in Ra led to chaotic yet symmetric vortex formations for low Pr fluids. The scaling of Nusselt number to the Rayleigh and Prandtl numbers, Nu = f(Ra,Pr) was investigated. Excellent agreement is obtained with previous numerical-benchmark solutions at Pr = 0.71 to ensure model accuracy. It is discovered that increasing Pr leads to increased Nu at high Ra-values and that Nu is most sensitive to Pr near Pr = 1. Results underline the importance of Pr upon the temperature and velocity fields and formation of vortices in differential heated cavities.   

In-situ measurement of contact angles for evaporating droplets using transmission interference fringe (TIF) technique

Recently we developed the transmission interference fringe (TIF) technique and verified it by experiment and simulation to show a good agreement. In this research, TIF technique is employed for the evaporating droplets on the plain glass substrate to show a good match compared with the contact angle by the sideview imaging technique. The contact angle is measured by measuring the fringe radius from the center to the edge ray from the dispersed beams on the screen away from the sample. The red laser (632.8 nm) is illuminated from the sample bottom at an incidence angle of 50 degrees with the transparent substrate. A total of three different volumes of droplets are used: 1, 2, and 3 ul, and the droplet sizes are close to the capillary length (~ 3 um). It shows that all droplets maintain the spherical profile with constant contact radius (CCR) mode. In TIF, contact angles can be easily detected in a simple configuration, not requiring a sophisticated microscope setup. The volume change rate is shown to match reasonably with the diffusion-limited model. TIF can effectively measure contact angles less than 1 degree in real-time, which is substantially important in ultra-small contact angle measurements in evaporation, condensation, and liquid film dynamics. However, the existing techniques are not easy. Furthermore, the top reflection interference fringe (TRIF) technique is formulated to verify that top reflection has the advantage of measuring the contact angles of droplets on the non-transparent substrates. With TRIF, it is expected that we can employ simultaneous measurement of contact angle or thin film thickness measurement from the top reflection interference fringe (TRIF) and the bottom reflection interference fringe (BRIF) or surface plasmon resonance (SPR) imaging technique for the ultra-small droplets or thin film with a few nanometer scales.   

Droplet Friction and Adhesion: The Solid-Liquid Amontons' Laws

The empirical laws of dry friction between two solid bodies date back to the work of Amontons in 1699 and are pre-dated by the work of Leonardo da Vinci. Fundamental to those laws are the concepts of static and kinetic coefficients of friction relating the pinning and sliding friction forces along a surface to the normal load force. Here, we show how surface free energy arguments can be used to derive and define coefficients of static and kinetic friction for three-phase contact lines and droplets on solids. We provide a droplet analogy of Amontons' first and second laws, but with the normal load force of a solid replaced by the normal component of the surface tension force of the liquid. In the static regime the coefficient of static friction is proportional to the contact angle hysteresis. In the kinetic regime the coefficient of kinetic friction is proportional to the difference in dynamic advancing and receding contact angles. Our Amontons' laws provide a direct link between droplet friction and work of adhesion. We show literature data can be fitted to provide numerical values for droplet-on-solid coefficients of friction.   

Hele-Shaw drop deformation in quadratic extensional flow

The four-roll mill and microfluidic experiments have often used linear extensional flow to study deformation, break-up and coalescence behavior. There, however, is no experimental evidence of higher order flows being used to study emulsion drops. Here, we consider the deformation of a Hele-Shaw drop exposed to a quadratic flow that forms inside a microfluidic hexagonal chamber. A Hele-Shaw drop was held at the center of the chamber and at the stagnation point of the flow ‘indefinitely’, using a computer-controlled six-port extensional flow device, and the dynamic deformation of the drop was recorded at the applied stresses. For quadratic Hele Shaw flow, in the limits of small deformation, the drop deforms into a triangle with round corners as opposed to an ellipse in a linear extensional flow. The deformation is calculated in the limit of small capillary number, and the theory matches with experiment. This is the new way of deforming drops on a quadratic platform experimentally, and may lead to measurements such as interfacial tension and breakup.

    

Sloshing Resonance of Acoustically Levitated Air-in-Liquid Compound Drops

An acoustically levitated liquid droplet containing an air pocket is driven into an out-of-phase azimuthal 'sloshing' resonance by a frequency modulation with modes n = 4 to 9. The inner and outer liquid-air interface waveforms are consistent with the standard Saffren model. The frequency spectrum resonance peaks and their harmonics are found to be a function of droplet size and resonance modes. Due to asymmetry, droplets with several tiny air pockets do not resonate in time with the cavitation. This research has major implications for the dynamics of cavitating drops.   

Does the growing liquid bridge of drops coalescing in a Hele-Shaw cell feel the no-slip wall?

When drops come into contact with each other, a liquid bridge is formed and grows rapidly during coalescence. When drop coalescence occurs in a narrow confinement such as Hele-Shaw cells filled with a viscous liquid, the temporal growth of the liquid bridge is expected to be affected by the no-slip surface of the cell through the ambient liquid moved by the bridge. To investigate the hypothesis, we captured the liquid bridge growth of corn syrup drops merging in mineral oil confined by the Hele-Shaw cell using high-speed video microscopy, and we measured the width of the liquid bridge using image processing. Then, we determined the bridge width value at which the bridge growth deviated from the scaling law for the initial phase of drop coalescence. It was observed that the determined bridge width values showed a sudden change as the viscosity ratio between the drop and surrounding liquid changed. A simple theoretical model was developed to explain the observed change and to show that the growing neck could feel the no-slip wall of the Hele-Shaw cell depending on the viscosity of the surrounding liquid.   

Influence of dye concentrations of internal reflection suppression and air entrainment dynamics of polymeric droplets.

In many studies that incorporate droplet impact imaging techniques such as total internal reflection microscopy (TIRM), dyes and particles are often used for visualization enhancement. For the bottom view techniques such as TIRM, which depends on grayscale intensity changes to directly measure the spatio-temporal shifts in narrow gaps, internal reflections induced by the collapsed top surface can be seen and negatively affect the image quality. In this study, we show that with concentrations of various dyes such as Methylene Blue reaches a limit at which the internal reflections are suppressed, yet the air film topology is also affected. We attribute the changes in the air entrainment dynamics to both optical and physicochemical processes.   

Shock-induced bag breakup of droplet

The spatiotemporal dynamics of shock droplet interaction are studied for a range of Weber numbers between 8 and 15. This regime includes bag breakup mode, which is critical in understanding secondary atomization of liquid droplets and designing the optimum length of combustion unit in a Scram-jet engine. The Shockwave of the desired strength is generated in the squared section shock tube. The syringe pump produces a droplet of 1.8 mm diameter, which is exposed to shockwave for secondary atomization. The morphology of the droplet is captured using the shadowgraphy and high-speed camera at 33,000 FPS. The bag regime presents interesting flow features and mainly develops at low Weber numbers. This regime shows high sensitivity toward surface tension compared to other droplet deformation regimes. Flow visualization results in ensuing instability on droplets, their motion and evolution of bag with time will be presented. It has been observed that the critical bag size increases with an increase in Weber number. The relationship between optimal bag length and corresponding Weber number will be discussed.   

Impact of a droplet on a circular superhydrophilic region surrounded by a superhydrophobic region

The spatial variation in the wettability of a surface can have a significant effect on the spreading and retraction behavior of an impacting droplet and hence the overall impact dynamics. Although composite surfaces have proven applications, there is a lack of understanding of droplet impact on surfaces with a sudden jump in wettability. Here, we study the behavior of a liquid drop impacting a composite surface having a superhydrophilic (SHL) spot surrounded by a superhydrophobic (SHB) region. We find that the droplet exhibits different regimes: no-splitting, jetting, and splashing, depending upon the spot size (b_s)  and Weber number (We). At a smaller b_s, the behavior shifts from the stable to the jetting regime and then to splashing regime, with increasing We. We find that by increasing the value of b_s, one can avoid the undesirable splashing and jetting regimes and attain a stable regime even at a higher We. Our study reveals that b_s has a significant influence on the maximum spreading diameter B_max at a smaller We but a negligible effect at a higher We. We show that the dominance of capillary energy at a smaller We, and viscous energy at a higher We underpin the phenomena. We employ an energy conservation approach to develop an analytical model to predict B_max  on a composite SHL-SHB surface, by considering the total energy of the system before the impact and at the maximum spread position. We find  K  = (Re^1/2/ We) emerges as a key parameter in the model that accurately predicts the experimentally measured B_max. Our study reveals the existence of an inertia-viscous dominated regime at smaller K and an inertia-capillary dominated regime at larger K . The outcome of our study may find applications in stable and precise positioning of impacting droplets.

    

Influence of the relative humidity on the impact of blood drops on solid substrates

Although the receding (retraction) of a drop after its impact on a solid substrate is an important phenomenon, it receives less attention than spreading. We investigate the retraction dynamics for a drop of whole human blood with different impact velocities on wooden floors that were either varnished or not. The effect of relative humidity and temperature are considered by carrying out extensive experiments under different surroundings conditions. The obtained results show that retraction is present on the varnished wooden floor (smooth substrate), in contrast to the unvarnished wooden floor (rough substrate). For the first one, the retraction occurred only at low relative humidity. This was explained by the decrease in the surface tension with relative humidity level. We will present our results on blood drops retraction after impact by discussing its dynamics and the influence of relative humidity. 

    

Water entry dynamics of hydrophobic spheres near floating debris on a deep liquid pool

We take the enduring topic of water entry further by documenting splash features generated by the impact of hydrophobic spheres with floating debris atop a deep liquid pool. Proximal interaction with floating debris is yet another means to manipulate splash dynamics. In this experimental study, we systematically investigate the fluid-structure interactions between floating debris and hydrophobic spheres for Froude numbers in the range of 21-74, and sphere-debris separation distance. Thus, we observe previously undocumented sub- and supersurface fluid interactions. Generally, hydrophobic spheres produce a radially expanding splash crown just above the free surface simultaneously with a deep seal cavity characterized by smooth cavity walls below the free surface, and a vertically-protruding Worthington jet following cavity collapse. The proximal presence of debris atop the free surface with respect to impacting spheres prohibit radial fluid expansion, yielding lopsided splash crowns, non-uniform air-entraining cavities, and in some cases, lateral sphere migration. Such observations augur well for fluid-structure interactions where splash and trajectory control are desirable.   

A law of reflection for skirting droplets: the scattering angle of skirting droplets that rebound from a flat wall

Without a driving force, droplets that impact a bath with a horizontal component of momentum can glide or skirt along the surface of a bath of the same fluid, traveling many droplet diameters before coalescing with the bath.  These droplets will rebound from the meniscus formed on objects that protrude through the bath surface.  We show here the behavior of droplets that rebound or reflect from the meniscus formed on a flat vertical wall.  A linear but asymmetric law of reflection is found and compared with a numerical model that assumes an exponential meniscus profile.   

Water entry dynamics of hydrophilic spheres through particle-laden free surfaces

We advance the seminal topic of water entry by documenting splash features generated by the impact of hydrophilic spheres through particle-laden free surfaces. Proximal interaction with small, buoyant particles is yet another means to manipulate splash dynamics. In this experimental study, we systematically investigate the fluid-structure interactions between floating particles and hydrophilic spheres for Froude numbers in the range of 20-94. Thus, we observe previously undocumented sub- and supersurface fluid interactions. Generally, hydrophilic spheres entering a liquid bath below the threshold of 8 m/s produce minimal fluid displacement and no cavity formation. The presence of floating particles atop the free surface with respect to impacting spheres promote flow separation, yielding a radially expanding splash crown just above the free surface, simultaneously with an air-entraining cavity into the body of the fluid, and a vertically-protruding Worthington jet following cavity collapse. The resulting splash metrics differ from those of purely hydrophobic spheres based on the size of the impacted particles. Such observations augur well for fluid-structure interactions where flow separation warrant control.   

An accessible droplet-on-demand device that combines 3D printing fabrication and commercially available nozzles

Droplet on Demand (DOD) devices are an important tool for generating single droplets of controlled size in various fluid studies.  DOD designs have grown to be simpler and more accessible in terms of fabrication and cost by using SLA (resin) 3D printing techniques (Ionkin and Harris, Rev. Sci. Inst. 89, 116103 (2018)).  The greatest limit to 3D printed DOD devices is in fabricating the nozzle with a sufficiently small orifice to produce smaller droplets.  This challenge is further exacerbated when trying to use more common filament 3D printers.  We show here a simple and effective modification to the DOD design that replaces the fabricated nozzle with two alternative nozzles readily available on the market: brass 3D printing nozzles and Luer Lock blunt-tip syringe needles.  Both alternatives are available in a range of orifice sizes extending below the range of previously fabricated nozzles, thus allowing for droplet sizes smaller than those reported from work with recent DOD devices. We further show that the range of droplet sizes from a single nozzle/tip is larger than previously achieved allowing for more efficient experimentation.   

Droplet impact on the meniscus of a bath near a vertical wall

For many years droplet impacts on a bath of the same fluid have been studied for which the droplet has a velocity component parallel to the surface of the bath.  These studies have explored a range of bath-surface geometries (e.g. oblique impacts on horizontal surfaces and vertical impacts on angled walls).  For low Weber numbers bouncing can occur, which in turn can lead to droplet gliding or skirting along the bath surface, owing to the asymmetry of the impact.  Here we impose a parallel velocity component by allowing a droplet to fall vertically onto the meniscus of a bath near the wall of its container. Using low viscosity silicone oil, we explore the ranges in Weber number and the distance between the impact and the wall to determine the regimes for droplet bouncing and skirting.   

Dynamics of Droplet Pinch-off: Stiff and Soft Emulsions

Red blood cells are responsible for the rheological behavior of blood, which is a complex non-Newtonian fluid. Patients with sickle cell anemia (SCA) have reduced deformability of their red blood cells, resulting in flow blockage through the blood vessels. It has been shown that the rigidity of red blood cells and the concentration of rigid cells significantly affect the shear viscosity of the blood. While extensional viscous effects influence the microcirculation, its influence on SCA has yet to be understood. Using a model fluid and a uniaxial extensional flow experimental set-up, we investigated the effect of particle rigidity on the extensional viscosity. We studied the pinch-off dynamics of a thread of a suspension consisting of a photocurable emulsion and found that emulsions transition from a yield stress to a viscoelastic response upon exposure to UV. Although the mixtures of soft emulsion droplets show self-similar thinning behavior similar to Newtonian fluids, the thinning of an emulsion thread follows an exponential decay upon droplet polymerization. The magnitude of the elongational viscosity and emulsion relaxation time, calculated from the time constant of exponential decay, depend on the degree of polymerization and concentration of rigid droplets in the emulsion.

    

Computational investigation of non-trivial breakup morphologies observed during critical secondary fragmentation

In our previous work on the threshold of Droplet fragmentation under impulsive acceleration, we discovered that in addition to the trivial backward bag breakup normally observed for water-air analogous droplet systems, other breakup morphologies such as forward bag and backward bag-plume may also be observed. These non-trivial morphologies form at critical Weber numbers only for specific combination of parameters dictating the fragmentation process: density ratio (ρ), and drop (Oh_d) and outside (Oh_o) Ohnesorge numbers. Full 3D Volume of Fluid based direct numerical simulations for two representative cases are performed to capture all the physics pertaining to the two fragmentation morphologies under consideration. Features such as forward pancake formation and flipping of bag orientation for forward bag breakup; generation of plume on the upstream pole and its unstable growth for backward bag-plume breakups; generation of non-axisymmetric instabilities in the inflated bag; and the corresponding internal flows are explored in detail. This allows us to propose physically consistent fragmentation mechanisms for forward and backward bag-plume breakups.   

Bridge equilibrium solutions connecting a liquid lens and a sessile drop

We study the deformation of an axisymmetric sessile drop of liquid A on a horizontal substrate when the surrounding liquid B partially covers it. Thus, the drop adopts the shape of a liquid bridge that connects the substrate with the interface between the surrounding liquid and the air (fluid C). By applying both the Neumann's law at the triple point where fluids A, B and C meet and the Young's law at the contact line formed by the interface between liquids A and B at the solid, we find set of six second order differential equations along with the corresponding twelve boundary conditions. Aside from the symmetry conditions, we also apply no--flow conditions at the wall of the cylindrical container of liquid B. We numerically solve the equations by determing the nine unknowns parameters of the system, and we find the shape of the three interfaces (AB, BC and AC). We find two solutions with a neck. One of the solutions has at least one inflection point, while the other one does not. By performing an energy analysis, we find that the former is more likely to be found in nature (less energetic). In order to assert if a neck rupture is likely to occur, we compare the bridge energies with that of the corresponding separate drops configuration (a sessile drop plus a liquid lens).   

Author not Attending

In this article, for the first time, superhydrophobic leaf surfaces of Cassiatora, Adiantum capillas veneris (ACV), and Bauhinia variegata Linn. Indian have been studied using contact angle measurements and water drop impact experiments. Superhydrophobicity of Adiantum capillas veneris (ACV) leaves are reported for the first time. Field Emission Scanning Electron Microscopy (FESEM) characterization was done on these leaves to understand their surface morphology and how it affects wettability. A special drying method was employed to ensure minimal surface damage under the ultra vacuum of FESEM. Contact angle measurements show that the receding contact angle is the lowest in the Cassiatora leaf because of the widely spaced low aspect ratio nano petals and bumpy blunt microtextures. Other plant surfaces showed a very low roll-off angle less than 10° and very low contact angle hysteresis due to the presence of higher aspect ratio and closely spaced nano textures, re-entrant morphologies, and ridge structures. Droplet impact studies on these surfaces performed at weber numbers of 9,30,62 and105 showed different behaviour due to different micro nano textures. Significant differences such as secondary droplet formation after droplet complete rebound and maximum droplet spreading at higher We no were observed and showed the influence of observed micro nano textured on these leaf surfaces.   

Energy Harvesting: a universal model for reverse electro-dialysis

Reverse-Electro-Dialysis (RED) allows us to tap into the naturally occurring concentration asymmetry between seawater and freshwater to generate sustainable electricity. This salt concentration asymmetry is abundant and self-sustained by way of the water cycle and thus presents a novel sustainable energy source that could revolutionize the global energy market. However, much of RED’s microscopic details are not thoroughly understood due to the complex geometry and material properties of the ion-selective membranes. This has led to the conventional reliance on trial-and-error for optimizing RED systems. To overcome these challenges, we derived a universal model for all RED systems that accounts for the key parameters of these systems. We found analytical expressions for the open-circuit electric voltage and the Ohmic resistance. Our model is confirmed by numerical simulations and existing analytic models and thus delineates the transport characteristics of RED systems. Notably, the physics of energy harvesting via RED has been elucidated, and our findings can be used to improve and expedite the design of any RED system using permselective materials for ion transport.

    

DNA migration and dispersion due to simultaneous shear flow and electrophoresis.

Simultaneous application of axial electric and flow fields through a microfluidic channel can focus DNA at the center or walls of the channel if electrophoresis causes the DNA to lead or lag the flow, respectively.  The transverse migration is driven by hydrodynamic interactions caused by the electric field acting on the charged DNA molecule which is stretched and aligned by the flow field. Additionally, thermal fluctuations in the configuration of the DNA molecule correspond to different, instantaneous electrohydrodynamic velocities, and these velocity fluctuations contribute to the effective polymer diffusivity.  This electrohydrodynamic dispersion is comparable with or exceeds diffusivity due to Brownian forces for electric field strengths commonly used in microfluidic devices. Here, we present a model for electrohydrodynamic migration that also accounts for dispersion.  Competition between the electrohydrodynamic migration and dispersion is shown to cause a nonmonotonic dependence of DNA focusing on the electric field.  Predictions of the model are in quantitative agreement with Brownian dynamics simulations and in qualitative agreement with experiments.

 

    

Siting Optimization of In-stream Hydrokinetic Turbines within Hydropower Tailrace Channels

Hydrokinetic energy is a renewable energy source currently attracting attention from funding agencies and developers across the United States. Recently, interest has been given to deploying hydrokinetic energy devices, also known as current energy converters (CECs), within tailrace channels downstream of conventional hydropower plants, where favorable siting and operational conditions exist. However, beyond the economic, environmental, and technical challenges, in-stream turbine deployment may reduce the hydraulic head available for upstream hydropower production, causing a reduction in hydropower generation. Under simplified assumptions, we present a one-dimensional momentum balance approach to analyze the backwater rise as a function of the hydraulic parameters and device characteristics. Model results were validated against three previous laboratory and field measurements at different scales. Based on the analysis, deploying a CEC device too close to the upstream dam would likely induce hydropower losses that outweigh the hydrokinetic gain, resulting in a negative net system power production. However, this can be avoided by deploying the turbine sufficiently downstream to let the backwater effect recovers to the upstream undisturbed water level. Combining the traditional backwater equation for open channel flows, an optimized siting distance can be computed to maximize the net power production while remaining within the tailrace boundaries to capitalize on its engineered channel, predictable outflows, and proximity to grid interconnection. This study presents the preliminary proof of concept for future designs of the hydrokinetic system in tailrace channels.   

A cheap and rugged field PIV system

To date, there is a lack of turbulence data in natural aquatic systems. Adapting particle image velocimetry (PIV) to field use will provide such data. We present a new field PIV system optimized for high portability and low cost. This system was tested in the Huron River in Ann Arbor, MI, at various locations along the river. All measurement sites were close to shore (with depths of less than 1.5 m), and sampled a variety of environments, including bulrush marsh, pebble bank, boulder field, and woody debris. Turbulence metrics significant to aquatic biota were captured. Velocity was validated through comparison with an acoustic doppler velocimeter (ADV) and an electromagnetic flowmeter, and had an average 4.8% difference from these methods. Turbulence kinetic energy was validated through comparison with an ADV, and fell within one standard error. Vorticity, eddy size, and eddy circulation were validated at a field site by repeating measurements taken using a different field PIV system. Each of these data sets fell within the range of the validation measurements. This novel, low-cost, highly portable system provides accurate turbulence measurement, therefore offering an accessible method of acquiring meaningful field turbulence data for a wide range of users.   

Experimental study on the calibration of flow distribution in water circulating channel for resistance and self-propulsion test of underwater robots

This study aimed to validate the speed at the boundary layer in the water circulating channel and examined the effect of surface flow acceleration. When the surface speed accelerator is operated, the distribution of the flow rate increases by 5-10% from the set flow rate near the inlet. The surface flow speed gradually decreases as it moves away from the inlet and reaches the desired flow speed.

Therefore, it was necessary to adjust the rate when conducting an experiment near the water-air surface in the water circulating. The flow velocity distribution at the boundary's vertical and horizontal directions is a significant parameter. In the future, it plans to make the acceleration effect of the flow speed through the particle image velocimetry.   

Vortex formation in the wake behind various surface-piercing cylinder configurations

The offshore structures were developed with the growth of discovering world resources in the oceans. The vortex formation in the wake of the various cylinder cross-section configurations requires more illustration. The purpose of this study is to analyze the multiphase flow effects over the different cylindrical structures using numerical simulations. The computational calculations were performed using a validated numerical model. The cylinder configurations were circular, streamlined foil, elliptical, and merged foils in the vertical position. The results of flow over cylinders were compared at different flow Fr up to 4 at calm water conditions. The examined flow was compared for the separation characteristics and the wake development behind the cylinders. The results demonstrated that the modification of the traditional circular cylinder produces different flow wake characteristics. In addition, the free surface significantly alternated the generated flow, especially with a low vertical submerged depth. These findings have significant implications for understanding the flow physics underwater and developing offshore applications. Offshore applications range from small-scale cylinders like cables to massive structures such as offshore platforms.   

Utilization of Schlieren Imaging to Investigate the Effect of Active Control on Multi-Aperture Rectangular Single Expansion Ramp Nozzle

Schlieren imaging is a well-known method to investigate and record a flow field’s structure. It is an invaluable tool to record the affect of an active control composed of micro-jets in the flow field of the shockwaves created by a Multi-Aperture Rectangular Single Expansion Ramp Nozzle (MARS). The Schlieren imaging system used is designed vertically around the MARS nozzle to accurately capture the developing shock structures. Previously, research was performed using a passive control which was proven to diminish a dominant frequency tone created from the instability at the interface of the core and bypass streams within the MARS. To expand and improve upon the past methods used, active control is implemented in the splitter plate trailing edge region, a region determined via simulation to be highly receptive to excitation. Schlieren imaging is used in order to record and understand the change resulting from the active control. By using this method, data from the active control can be compared to previous experiments that used a passive control as well as the uncontrolled case. The comparison of the data from the two control systems is extremely important to determine the effect of an active control on the restriction of the dominant frequency that arises in the flow field.

    

Drag and heat transfer in azimuthally oscillating turbulent pipe flow

Experiments are preformed to investigate the effects of transverse momentum injection on the drag and heat transfer of fully developed turbulent pipe flow. A driving system consisting of a crank-slider mechanism is employed to oscillate the pipe azimuthally, so that the spanwise wall velocity is varying sinusoidally in time. Pressure and temperature gradients along the oscillating pipe are measured to determine drag and heat transfer, respectively, and the results are compared to the power consumption of the driving system to characterize the overall energy efficiency. Results will be discussed for two friction Reynolds numbers of approximately 4000 and 8000.

    

Surfactant-induced instability of foam in a radial Hele-Shaw cell

Foam, i.e., dispersion of gas bubbles in a liquid, is at the center of our daily activities, from medical applications to cleaning purposes, and enhanced oil production. In contrast to single-phase gas, foam demonstrates high viscosity which makes it a suitable agent to displace oil from an underground oil reservoir. In this study, we investigate the influence of surfactant dilution on the dynamical behavior of foam movement in a radial Hele-Shaw cell. To generate foam, the co-injection of nitrogen and Sodium Olefin Sulfonate (17 times above critical micelle concentration) into a micro-model is considered. The generated foam with the constant gas fraction (foam quality) is injected directly from the micro-model to the Hele-Shaw cell. The displacement is monitored when the Hele-Shaw cell is filled with (i) water and (ii) a surfactant solution that has the same surfactant concentration as the injected foam. Outcomes show that the displacement process remains stable in surfactant-saturated Hele-Shaw cell. The displacement of DI-water by foam, however, becomes unstable with the advancing fingers of foam into the DI-water. Measurement of the size of bubbles during the injection illustrates that the foam destabilization mechanisms, comprising coarsening, liquid drainage, and coalescence are not active. As the size of bubbles does not change over time, the branching pattern of foam in water-filled Hele-Shaw is not attributed to film rupturing. For one set of experiments, we add dye in the DI-water from which we can monitor the growth of the dilution zone, i.e., water mixing with the liquid fraction of foam. It is hypothesized that the water diffusion into foam reduces the surfactant concentration which enhances the capillary forces and consequently promotes the side branching pattern. Consistently, an increase in the liquid fraction of foam suppresses the fingers as the mixing zone becomes less diluted due to the larger volume fraction of surfactant solution in foam.

    

Miscible Rayleigh-Taylor Instability Experiments on the Linear Induction Motor (LIM) Drop Tower

Incompressible, miscible experiments on the Rayleigh-Taylor Instability (RTI) using Planar Laser Induced Fluorescence (PLIF) imaging are presented. A vertical tower guides a test sled that is accelerated downward using linear induction motors. Experimental liquid pairs are prepared and placed into a test chamber attached to the sled. The sled is accelerated at a rate of approximately 10.5g. Upon acceleration, the stratified initially stable fluid pair within the chamber becomes unstable causing the development of RTI. The resulting RTI is imaged using PLIF by seeding the heavier fluid with fluorescein dye that is illuminated by a 445nm wavelength laser sheet. Variation of the indices of refraction (IOR) of the two fluids occurs during mixing, producing image blurriness and negatively affecting PLIF Imaging. This IOR variation is minimized by modeling the IOR of the fluids as a nonlinear fluid property, allowing preparation of optimized fluid pairs. Both unforced and forced experiments are performed. Forced experiments are conducted to produce initial perturbations and capture more of the RTI growth within the experimental observation window. Measurements of the resulting mixing layer growth and growth rates will be presented.   

Minimizing perturbation growth along interfaces accelerated by large pressure waves

Large pressure waves passing through an interface separating a heavy and light fluid can cause hydrodynamically unstable perturbations to grow. Minimizing this growth is crucial for achieving the high-pressure regimes required for a sustained nuclear reaction in inertial confinement fusion (ICF). In ICF, the laser-driven pressure waves typically consist of a shock front followed by a decompression wave caused by the termination of the laser pulse, causing the interface to evolve under the influence of a complex combination of Richtmyer-Meshkov and Rayleigh-Taylor instability growth. Our objective is to find initial pressure profiles that minimize perturbation growth for a given interface. As a model pressure wave, we consider a shock front being overtaken by a rarefaction wave. Our analysis indicates that perturbation growth can be minimized if the shock-induced phase inversion, which initially causes the perturbations to decrease in amplitude, is counteracted by the rarefaction-induced growth. We further identify the relationship between the shock strength, rarefaction strength, and rarefaction length that minimizes growth. Our analysis, which is grounded in analytical techniques from gas dynamics, is verified by comparison to high-order-accurate numerical simulations.   

Global sensitivity explaining atmospheric shear layer transition

As small vortices appearing within a mixing layer in the atmosphere often shake airplanes dangerously, a better understanding of its transition process is necessary to improve current turbulence forecasts provided by meteorological organizations. There is little meteorological knowledge of that phenomenon because it often happens above the atmospheric boundary layer where weather forecasting models are solved on coarse grids and intensive turbulence observations are usually difficult. We investigated a secondary rib-like structure growing on the primary Kelvin-Helmholtz (KH) vortices in a realistic atmospheric shear profile, which was simulated by a weather forecasting model. By using a three-dimensional Euler computation, we obtained global sensitivities to the primary vortex by measuring kinetic energies of superimposed sinusoidal velocity and vortex disturbances. We concluded the growth of the rib-like structures was explained by the vortex stretching mechanism due to a strong strain between two KH vortices, which was not obtained by the conventional linear stability analysis.   

Fluid Dynamics Monitoring using Internet of Things (IoT) Technology

Fluid dynamics is essential for understanding many environmental challenges including but not limited to carbon footprint and water quality issues. Pollution in both air and water can trigger many respiratory and neurological disorders which can be life threatening. Pollution can cause various diseases not only to humans, but it also can affect animals and other species, which in turn affects the entire life cycle ecosystem. In this study, we use an array of  high precision embedded sensors to record real-time data that can be used to assess the indoor air and water quality. These sensors are integrated in a cloud-based platform using internet of things (IoT) technology for accurate monitoring of several fluid quality indicators. For example, sensors for measuring water turbidity, pH, conductivity, biological oxygen demand, suspended solids, carbon dioxide, ozone and organic volatile compounds and several others are used to build a custom-based monitoring station. The collected data from this environmental indoor station can be used to help and educate the public to take enough precautions to avoid chronic and acute diseases related to air and water pollution.   

Drag reduction in turbulent boundary layers using superhydrophobic surfaces

We studied the use of superhydrophobic (SH) surfaces for the frictional drag reduction in turbulent boundary layers. First, SH surfaces were manufactured and characterized in laboratory. Then, the reduction of frictional drag over these SH surfaces in turbulent boundary layers at the Reynolds number ranging from 4.0 × 10⁵ to 10 × 10⁵ was measured in a water channel using a dedicated, high-resolution force measurement system. The near-wall velocity field was measured with a particle image velocimetry system. Results revealed that, regardless of the Reynold number, the SH surface can achieve significant drag reduction if a homogeneous air film is able to form. However, we also observed that, as time advanced, air bubbles started to form and accumulate on the trailing edge of the SH surface and gradually broke the uniformity of the air film. When the inhomogeneous gas film formed, the SH surface exhibited less drag reduction and even drag increase. This evolution of air firm has been systematically studied in our experiments. In addition, high-fidelity computational fluid dynamics (CFD) simulations were also conducted to provide more detailed information on the frictional drag experienced by the SH surface.   

Exploring the inequality of evaporation and condensation coefficients based on ISS experiments

Phase change is a complex problem, and kinetic theory has been used to model it for over a decade. However, due to unknown values of evaporation and condensation coefficients, the applicability of the kinetic theory is still limited. These coefficients are necessary inputs and denote the fraction of molecules that undergo phase change. For the sake of simplicity and closure, the evaporation and condensation coefficients are often assumed to be equal (and often called an "accommodation" coefficient), The current study aims to investigate this assumption, using the Constrained Vapor Bubble (CVB) experiment data from the ISS. Interferometric image analysis of the CVB data enables a complete reconstruction of the liquid-vapor interface. Thermal transport is modeled using the temperature profiles obtained from the CVB dataset. A multi-scale phase change model is developed to calculate the local phase change flux at all points across the 3D interface using a unique "active surface" meshing technique. The model accounts for the interfacial curvature, disjoining pressure, and thermo-capillary effects in the contact line region. The CVB cell being a closed system allows for an additional governing equation: the surface-integral of the phase change flux over the entire interface is zero. The additional equation enables the decoupling of the evaporation and condensation coefficients and indicates a potential inequality.   

An Analysis of the Effectiveness of Flow Conditioners in Reducing Turbulent Kinetic Energy in Pulsatile Pipe Flow

Turbulent flow produces significant fluctuations in fluid dynamic properties including volumetric flow rate, wall shear stress, and pressure. Moreover, in benchtop experimental setups, the use of equipment such as gear pumps can induce undesired and unpredictable flow disturbances that can lead to inaccurate or inconsistent flow measurements. Thus, for such experiments, proper flow conditioning is critical, to ensure measurements are highly repeatable and initial conditions are consistent across test cases. While significant research has been conducted on flow conditioning in large-scale environments, few studies have investigated optimal conditioners for use in small-diameter (D < 2 inches) and pulsatile flows, often needed for in vitro hemodynamic studies. In this work, we compare the performance of large-diameter flow conditioners (the Laws plate, Etoile, and tube bundle) at reducing the turbulent kinetic energy (TKE) in small-diameter, physiologically-relevant flows. We use both planar particle image velocimetry (PIV) and computational fluid dynamic (CFD) to evaluate the TKE upstream, within, and at several locations downstream of the flow conditioners. Reduction of TKE as well as the induced pressure drop were used to evaluate the efficacy of each flow conditioner tested.   

Enhancement of capillary-driven viscous fluids flow under electric field effect

Transportation of fluids through capillaries by micropumps has been stimulated by considerable interest in micro-electro-mechanical systems (MEMS) technologies, microfluidic devices, and biomedical engineering. Enhancing the sluggish underfill flow of high viscous fluids is crucial for device commercialization, therefore improving the slow speed of such viscous fluids is one of the primary concerns. Along with laboratory investigations, deeper elucidation of the viscous fluid dynamics between parallel plates involves the continuous requirement for more elegant numerical simulations and mathematical models. In this study, it was investigated how diverse viscous fluids flow between parallel plates, driven by the combined effects of capillary and electric potential can be used to achieve speedy filling. An electric potential actuation induced a 40% faster filling of viscous fluids between parallel plates when compared to a basic capillary flow. The theoretical model was also used to study the dynamics of viscous fluids flow and the results were found to be consistent with the experimental values. By applying an electric potential, the filling time was significantly shortened by 30% for 3% NaCl in glycerol. This study can be widely used to enhance the viscous fluids flow speed for underfill and MEMS applications.   

Urban Air Mobility Wind Collection: collecting wind and atmospheric data to improve safety and efficiency of low-altitude UAS operations, specifically in urban environments.

UAS technologies are becoming more widely utilized in civil and commercial fields and military applications. Amazon's drone delivery service and Boeing's eVTOL air taxi are some examples of this. With small urban UAS applications becoming common, infrastructure, such as UAS traffic management for low-altitude airspace management and monitoring, is being developed. Safety and efficiency of UAS operations are strongly impacted by low-altitude wind, such as gusts around buildings. Gusts can negatively affect pilot operations, reduce flight time, and cause damage to the UAS system. For this project, a fleet of specialized UAS quadrotors will collect local wind data around buildings and urban environments and transmit the data in real time to ground stations in said urban environment to help in creating a real-time suggested flight path. Wind measurements will be taken around buildings on the Oklahoma State University campus, specifically the Kerr-Drummond buildings and Boone Pickens Stadium. One UAS quadrotor with a Trisonica Ultrasonic Anemometer to measure wind speed and direction was utilized in initial testing phases around such urban environments. Once wind data can be accurately and quickly collected around urban environments, more UAS systems with the data collection technology will be built and implemented. After several UAS systems can capture wind speed in multiple locations, a system will be built to send the wind data from the UAS quadrotors to a ground station in real-time.   

Arctic Sea Ice Drift Predicted Using Machine Learning

The movement of sea ice is influenced by a number of factors, from winds to ocean currents. As climate change continues to advance, sea ice drift in the Arctic is a key parameter to understanding the effects of rising temperatures in the region and around the world. Research has shown that the Arctic and the Antarctic are warming faster than the rest of the planet, which raises questions regarding climate justice, as most of the carbon emissions causing anthropogenic climate change are produced in other regions. To analyze this impact, we employ artificial intelligence to predict sea ice drift velocity based on external features. Machine learning is the process of computers gaining insights by seeing patterns and correlating large quantities of data. Using external parameters, including wind speed, and drift velocity ground truth as the inputs of the model, we train multiple different architectures and compare the results. In particular, we experiment with a convolutional neural network (CNN), a random forest (RF), and a support vector machine (SVM). We also experiment with various model specifications. The ultimate aim is to achieve a greater understanding of the Arctic's response to climate change.   

Quasi-geostrophic convective turbulence at large Rayleigh number

An asymptotically reduced model for Rayleigh-Bènard convection in a planar geometry is numerically investigated up to reduced Rayleigh numbers (Ra_r = Ek^4/3Ra, where Ek is the Ekman number and Ra is the depth-scaled Rayleigh number) of 280. We carry out a suite of simulations wherein the barotropic (depth averaged) components of the flow are zeroed out at each time-step in order to remove the effects of domain-sized vortices. The results are compared to cases where the barotropic component is not removed. Without the large-scale vortices, a linear scaling between the reduced Reynolds number Re_r = Ek^1/3Re and Ra_r is reported as well as a Ra_r^3/2 scaling for the Nusselt number. In addition, it is found that the Taylor microscale is approximately constant with increasing Ra_r, despite the increase in global Re_r. The kinetic energy spectra are not found to follow a Kolmogorov -5/3 law, which suggests the absence of a traditional inertial range. We interpret these results as evidence that energy in rapidly rotating convection is generated at small (viscous) length scales; kinetic energy cascades to both larger and smaller length scales, the former of which is responsible for the appearance of large scale vortices.   

Spreading Cohesive Powders

In many additive manufacturing processes, the 3D structure is built via solidification of sequential layers of powder. To create each layer, powder is spread over a horizontal surface.  The layer properties are determined by the machine geometry, process conditions and powder properties. Discrete element modelling (DEM)  can be used ro simulate the system.   The inclusion of cohesive forces in DEM powders increases the complexity of the contact models and particle stiffness scaling is often used to decrease simulation time.  This study uses an efficient GPU code to test a new scaling methodology and apply it to cohesive particle spreading.  In this work we demonstrate the validity of proposed model, both for powder packing and for spreading of fine powders before the effect of powder cohesivity on spreading is then investigated. A wide range of cohesivity, quantified via Bond number, was studied, significant differences in the structure of the spread layer were noted and several regimes identified. In particular, significant clustering, and less uniform spreading, was observed with lower cohesivity, flowable, powders. The mechanisms behind the observations are discussed and a dimensionless inertial number proposed to help interpret the phenomena observed.

    

Granular Dam Break: The role of Particle Shape.

We performed Discrete Element Method (DEM) simulations and experiments in a rectangular setup to investigate the influence of particle shape on the collapse of a granular pile. We observed that particle shape does not have a strong influence on the final run-out distance and all particle shapes followed the same power-laws as were proposed in previous studies. We found out that the maximum frontal velocity initially increases linearly with increasing aspect ratio, but then starts to bend towards a constant value. We suggest that the behavior observed maximum frontal velocity is associated with the Rayleigh-Janssen effect where the side walls shield the pressure exerted by the mass added on the top of the pile through wall friction. We observed that the time evolution of energy takes place in three stages. In the initial stage, the total initial potential energy stored in the pile partly converts into vertical kinetic energy, which, in the second stage, partly converts in horizontal kinetic energy until, in the last stage, basal friction dissipates the kinetic energy of the particles and halt the collapse. We also observed that the efficiency of the conversion of vertical kinetic energy into horizontal kinetic energy tends to decrease with increasing aspect ratio.   

Author not Attending

We consider the problem of mass transfer between liquid steel and slag during secondary metallurgy, a phenomenon that governs the adjustment of liquid steel composition. We study this phenomenon with a reduced scale water experiment which reproduces the dynamics seen in an argon-gas bottom-blown ladle. We develop correlations to characterize the hydrodynamics and mass transfer rate for a given Froude number based on a combination of data generated from Direct Numerical Simulations, experimental data, and data found in literature. For the hydrodynamics we develop a correlation for the evolution of the open eye as a function of air flow rate by considering a combination of elevation theory and viscous entrainment theory. We simulate the mass transfer of thymol between water and oil using a subgrid-scale model to resolve the thin concentration boundary layer present at the interface. We demonstrate relative agreement between the experimental and simulated Sherwood number and build a correlation for the Sherwood number as a function of the Froude number.   

Not Participating

Pattern formation and bubble size of magnetohydrodynamic (MHD) bubbly jet driven by the Lorentz force is presented experimentally. The Lorentz force generated by perpendicularly placed magnetic field and electric field displacing conductive saltwater, chemically produced gases (hydrogen, oxygen and chlorine) forms a typical two-phase MHD jet flows. Taking advantage of the bright gases, the emergence of bubbly jet and bubbles are analyzed. Based on the control parameters, such as magnetic field strength, input current strength, geometry of the experimental apparatus, and fluid properties, a Lorentz-force based Reynolds number $Re_L$ is used to categorize the flow regimes and size distributions of bubbles.   

Modeling a Low-Re Passive Microfluidic Mixer for Bioprinting Applications

Mixing at microfluidic scales is challenging due to the dominance of diffusion in laminar flows. Of the microfluidic solutions that have been designed and studied to increase mixing within microfluidic mixers, passive mixers provide advantages such as reliability, manufacturability, and straightforward integration into pressure-driven systems. In particular, two-layer crossing-channel micromixers (TLCCMs) induce chaotic advection at low Reynolds numbers (Re), resulting in high mixing efficiencies when compared with other passive mixer designs. Previous work has characterized this class of micromixer down to Reynolds numbers on the order of 0.01, but applications such as bioprinting often deal with lower Re ranges due to the high viscosities of bio-inks. In this work, we model and evaluate a TLCCM at these lower Re ranges using finite volume numerical methods. We determine mixing efficiency over a range of low-Re flows, for a series of TLCCM geometries, and as a function of the number of consecutive mixing units. The results from these models will broaden the current design space for TLCCMs and serve as a foundation for the eventual implementation of the microfluidic mixer into a bioprinting system.   

Dynamics of the capillary rise in tilted Taylor-Hauksbee cells

In this work, we theoretically and experimentally study the issue of the spontaneous capillary rise of a viscous liquid in wedge-shaped tilted cells, with very short angles of aperture, α. We provide the equilibrium profiles y_se and, by means of the Reynolds lubrication equations, we find the time-dependent profiles and the dynamic evolution of the meniscus close to the edge of the wedge, as a function of time, which follow power laws of the form y_s ∼ t^1/3. Experiments performed at various inclinations are consistent with our theoretical results.   

Modeling the Phase Transition Between Localized and Extended Deposition of Particles in Porous Media

Improving filtration efficiencies and lifetime of filters relies on an accurate prediction of the progression of clogging in porous media. More specifically, we try to explore how system properties, such as the applied pressure and network structure, affect the deposition of colloidal particles in disordered packings of glass beads. Past experiments have shown that depending on the applied pressure across such systems, we may expect either localized deposition (at lower pressure) or extended deposition (at higher pressure). We develop a mathematical model and use agent-based simulations to capture the results previously observed in experiments. To do so, we apply previously formulated deposition and erosion laws that lead to these two different behaviors and discuss the existence of a phase transition theoretically.   

An Additively Manufactured Small Footprint Wind Tunnel for Wall Jet and Particle Scavenging Studies

Powder bed fusion is a fast-growing additive manufacture method. However, the quality of produced metal parts is often challenged by particle spatter in the build chamber. To remove the spatter particles, it is important to gain further understanding of particle scavenging process. 

The focus of this paper is on the design and the build of a wall jet wind tunnel as well as experimental test data of particle scavenging inside the wind tunnel. The wind tunnel is additively build and has a co-flow configuration where a small-scale open loop suction tunnel creates uniform free stream, and a blower generates a plane wall jet through a rectangular slot. The wind tunnel is also equipped with adjustable height build plate on which metal power can be spread, variable speed axial fan, diffuser, honeycomb, perforated plates, and parallel air filters. The build-in flexibility of the wind tunnel allows testing various of different cross-flow configurations. 

A high-speed camera captures particle motion across a wide range of test condition for a number of metal powers used in additive manufacturing including aluminum alloy AlSi10Mg, nickel-based superalloy Inconel 718, titanium alloy Ti-6Al-4V, steel alloy 4340, and 316L stainless steel. Particle tracking velocimetry (PTV) and a high frequency hot-wire air flow meter are coupled to measure time-averaged flow field data.    

Optical investigation of gas-liquid interaction inside a counterflow atomizer

Complex gas-liquid interaction in the minichannel passages of an internal mixing counterflow atomizer was optically investigated using quartz capillary tubing and a diffused pulsed laser backlight. Image binarization techniques quantified gas penetration length upstream of the injection plane, and results compared to recent flow-blurring literature. Downstream, imaging also tracked the development of an annular flow in the first ten diameters of discharge tube, estimating film surface speed via cross-correlation of image pairs. Knowledge of internal and exit orifice flow conditions contributes to understanding the physical operation and droplet size scaling behavior of counterflow atomizers.   

Author not Attending

We have studied the atomization of a radially expanding thin liquid sheet formed by impingement of a liquid jet on a circular impactor. The sheet breaks from the rim into droplets. These experiments are performed in ambient conditions. Disturbances were introduced at the impingement point to investigate the effect of external forcing frequency and amplitude on the sheet break-up length and resulting drop size distribution. We observed that the sheet becomes more unstable resulting in a decrease in sheet break-up length and average drop size on increasing either the forcing frequency or the amplitude. This also resulted in a narrower drop size distribution.

    

Modeling water jet flow inside cylindrical chamber

Using CFD simulations, we investigate the flow of water injected from a nozzle tangent to the wall of a cylindrical chamber to generate a continuous water film covering the inner chamber surface. Our computational model is based an algebraic volume of fluids (VoF) method that considers the relative velocity in two sides of interface together with the piecewise linear interface calculation (PLIC) method to assure the boundedness and conservativeness of the interface. Two turbulence models are considered to model the effect of turbulence, k-omega SST and elliptic blending Reynolds stresses model (EBRSM). The results show that both turbulent models can correctly predict the attached water film. We validate the model and employ it to probe the effects of jet flow rate and the nozzle configuration on the water distribution on the surface of the chamber and the evolution of the water film thickness. The results can be useful for improving heat transfer in technological processes.

    

Implementation of Diffusion and Reaction Mechanisms for Reactive Ejecta Simulations

When a metal with a perturbed surface is strongly shocked, the surface generates micron-sized particles from the surface upon release of the shock wave as a limiting case of the Richtmyer-Meshkov instability (RMI). The creation, evolution and transport of these particles, commonly known as ejecta, is an active area of research which is important for the understanding of materials in shocked environments. This project intends to investigate the scenario of metal ejecta transport in a carrier gas with which the metal will undergo a chemical reaction. The project is being executed in the Lagrangian hydrocode multi-physics solver FLAG, and research is being done on implementing mass diffusion and reaction mechanisms to simulate these physics. The mechanisms will allow for the analysis of the development of a reacted metal layer (e.g., hydride or oxide), internal stresses within the material, and deformation of the particle under extreme conditions. Future work on developing additional components for fully resolved simulations of reactive ejecta particles, after verification of the current mechanisms, is also outlined.

    

A geometric PIC discretization of Lie-Poisson brackets

Non-dissipative models of fluids are known to possess a Lie-Poisson Hamiltonian structure. In discretizing such brackets, one encounters a closure problem: given a finite representation of the fields, it is usually not the case that the dynamic evolution of those fields is prescribed only in terms of that finite dataset. Particle based representations (e.g. point-vortex methods) for the 2D Euler equations circumvent this difficulty, but suffer from slow convergence and an unwieldy distributional representation of the field. We present the "dual PIC" method for the 2D Euler equations, a method based on PIC methods from plasma physics but with a dual Galerkin representation in addition to the particle based representation. The two representations are related to each other through an L2 projection. Moreover, the error in this projection is conserved as a Casimir invariant of the flow. While this method is presented in the context of the 2D Euler equations, it holds promise as a general method of Lie-Poisson systems   

Using covariant Lyapunov vectors to explore chaotic dynamics with long-range spatial coupling

Exciting progress has been made using powerful ideas from dynamical systems theory to describe chaotic fluid dynamics as a trajectory through an infinite dimensional state space. The covariant Lyapunov vectors (CLVs) describe the magnitude and direction of the growth or decay of small perturbations about the nonlinear trajectory which can provide new physical insights. An important aspect of fluid systems that is often not present in the models used to explore CLVs is the local and long range spatial coupling. For example, the nonlinear convective term in fluid systems can lead to large-scale mean flows in addition to localized coupling. Computing the CLVs for a fluid system is computationally intensive which makes it difficult to study fundamental questions such as this. Instead, we use 1D and 2D lattices of coupled maps with local and long-range spatial couplings which are chosen with fluid systems in mind. We explore these dynamics using the CLVs to gain new insights relevant to chaotic fluid dynamics.   

Data-driven dynamics of Kolmogorov flow on the inertial manifold

The high dimensionality of the Navier-Stokes equations (NSEs) hinders controllability of the flows and increases computational costs which in turn motivates the need to find reduced order models (ROMs). In this work we learn high precision low-dimensional models for the NSEs, specifically for two-dimensional Kolmogorov flow. We consider the onset of where chaotic dynamics occurs which exhibits quiescent dynamics that travel near the vicinity of relative periodic orbits followed by bursting events. An undercomplete autoencoder is used to find the inertial manifold dimension and a dense neural network is then used to time map in the reduced space. At a dimension of five, as opposed to the full state dimension of 1024, we see agreement at short times where the predicted trajectory travels close to the true one for approximately two Lyapunov times. Long time statistics related to input power, dissipation, and hibernating/bursting time fractions are also captured. The ROM is also used to predict the phase evolution in the x-direction as well as bursting events, showing agreement and high accuracy at a dimension of five.

    

Inductive bias and information fusion with concatenated neural networks

Although data-driven modeling holds promise in many applications by lowering the computational burden, training deep learning models needs a huge amount of data. This big data might not be always available for scientific problems and may lead to poorly generalizable data-driven models. Exploiting prior knowledge about the problem, this study explores a physics-guided machine learning approach to build more tailored, effective, and efficient surrogate models. For our analysis, we focus on the development of predictive models for turbulent boundary layer flows over a flat plate. In particular, we combine the power-law velocity profile (low-fidelity model) with the noisy data obtained either from experiments or computational fluid dynamics simulations (high-fidelity models) through a concatenated neural network. We illustrate how the prior knowledge from the low-fidelity model results in reducing uncertainties associated with deep learning models applied to boundary layer flow prediction problems. We demonstrate that the proposed inductive bias produces physically consistent models that attempt to achieve better generalization than data-driven models obtained purely based on data.   

Predicting parametric spatiotemporal dynamics by multi-resolution PDE structure-preserved deep learning

Although recent advances in deep learning (DL) have shown a great promise for learning physics exhibiting complex spatiotemporal dynamics, the high training cost, unsatisfied extrapolability, and poor generalizability in out-of-sample regimes significantly limit their applications in science/engineering problems. A more promising way is to leverage physical prior to develop physics-informed deep learning (PiDL) frameworks. In most existing PiDL works, e.g., physics- informed neural networks, the physics is mainly utilized to regularize neural network training by incorporating governing equations into the loss function. In this work, we propose a new direction to leverage physics prior knowledge by baking the mathematical structures of governing equations into the neural network architecture design. In particular, we develop a novel PDE-preserved neural network (PPNN) for rapidly predicting parametric spatiotemporal dynamics, given the governing PDEs are (partially) known. The discretized PDE structures are preserved in PPNN as convolutional residual network blocks, which are formulated in a multi-resolution setting. The effectiveness and merit of the proposed methods have been demonstrated over a handful of spatiotemporal dynamical systems governed by unsteady PDEs.   

Contribution of turbulence motions to turbulent heat transfer in polymer drag-reduced flows

We examined turbulent heat transfer in a polymer drag-reduced flow by analyzing the DNS database for viscoelastic channel flows (Re_τ = 125 and Pr = 5) of a FENE-P fluid. The viscoelastic flows with drag reduction rates of 15%, 34%, and 52% were examined as well as the Newtonian flow. The turbulence structures associated with the turbulent heat flux were investigated through the three-dimensional joint probability density function (JPDF) of temperature fluctuations (θ') and streamwise and wall-normal velocity fluctuations (uˆ and vˆ). The distribution of the JPDF revealed a higher correlation between u′ and θ′ in drag-reduced flow than in Newtonian flow, which was shown to be mainly attributed to the enhanced low-speed streaks. The weighted integral of the JPDF also showed that the contribution of the turbulent motion to the Reynolds shear stress is approximately the same as that to the turbulent heat flux in both Newtonian and drag-reduced flows.   

Dye Effects on Droplet Pinch-Off Dynamics

The degradation of irradiated polymers such as aqueous Poly(ethylene oxide) (PEO) when in contact with various dyes potentially alters the molecular structure of the polymer. Dripping-onto-substrate experimental method was used to determine the variable extensional relaxation time and extensional viscosity of several polymer-dye solution pairs irradiated by a white light source. First, the pinch-off dynamics of aqueous PEO was validated. Second, various concentrations of cationic and anionic dyes were added to fixed concentrations of PEO. Third, the solutions were irradiated for 24 hours and tested again. The experimental results demonstrated a clear difference in the elasto-capillary regime of the neck-thinning dynamics from the polymer solution alone versus with the added dye and irradiation. 

    

Experimental evidence of the existence of optimal injected steam mass flow rates to get the maximal water gravity drainage driven into steam chambers of different shapes

Based on a simple model theoretical, in this work we experimentally study the problem of gravity water drainage due to continuous steam injection into elliptical or circular porous chambers made of glass beads and embedded in metallic, quasi-2D, massive cold slabs. These configurations mimic the steam condensation, for a given time period, during the growth stage of the steam-assisted gravity drainage (SAGD) process, a method used in the recovery extra-heavy oil from homogeneous reservoirs. Our experiments validate the prediction of the theoretical model regarding the existence of an optimal injected steam mass flow rate, per unit length, φ', depending on the chamber shape, to achieve a maximum recovery of a condensate. We show that there are different and unique recovery factors, depending on the chamber's shape, when measured as the percentage of the mass of water recovered with respect to the injected steam mass. Our results can be extended to actual oil-saturated reservoirs because the model involves the formation of a film of condensates close to the chamber edge that allows for gravity drainage of a water/oil emulsion into the recovery well.   

A Continuous-Time Random Walk Approach for Upscaling Anomalous Transport in Vesicular Porous Media

The emergence of non-Fickian transport in heterogeneous fractured porous media has been widely investigated in the literature. These studies typically report an early breakthrough and long tailing at late times mainly as a result of the formation of preferential pathways of fluid within the heterogeneous porous medium. However, transport in vesicular porous media with multi-modal pore size distributions, consisting of a porous matrix and fluid-filled cavities is an ongoing area of research.

The anomalous transport in a vesicular media originates from the occurrence of non-Gaussian and multiscale velocity distributions caused by local flow instabilities, vorticities, and stagnant zones. These local features of fluid flow result in a significantly different velocity distribution for the tracer compared to the fluid. We simulate the inert tracer transport in a porous cell problem with one cavity located at the center. Main features of transport, including the tracer time transitions, are extracted within a continuous-time random walk (CTRW) framework. We use breakthrough curves to derive the upscaled CTRW memory term for a vesicular media with cavities of high degree disorder. We quantify the effect of Peclet number, cavity shape, and cavity volume fraction on the transport process.   

Study of the Onsager's-reciprocity-principle-consistent approach used in the derivation of higher-order transport equations

Failure of the Navier-Stokes equations leads to utilizing the Chapman-Enskog and Grad moment method to propose higher-order transport equations to extend the applicability of continuum formulations into the continuum-transition regime. However, these equations based on only mathematical principles failed to cover the whole transition regime of flow due to the non-consideration of the non-equilibrium thermodynamics in their derivation process. That is why the derivation of higher-order transport equations is still an essential topic in the research field. In the present work, we propose to incorporate the Onsager's reciprocity principle, a cornerstone of linear irreversible thermodynamics, into our derivation process. We focus on the derivation procedure of higher-order transport starting from the Boltzmann equation of the Kinetic theory. Two different relaxation times have been used to ensure the correct value of  Prandtl number. For their closure, the derivation procedure involves the evaluation of unknown higher-order tensors present in the evolution equations for stress tensor and heat flux vector and constitutive expressions for stress and heat flux vector for the Grad-like and Burnett-like equations, respectively. Using this approach, we have proposed new higher-order transport equations known as OBurnet and O13  equations. These O13 equations are shown to be unconditionally stable for any wavelength and frequency and consistent with Onsager's symmetry principle and H-theorem. The OBurnett are reported to be stable for any disturbances and produce smooth shock structures, showing the existence of heteroclinic trajectory and positive entropy generation inside the shock at all Mach numbers. Therefore,   transport equations obtained using this approach are expected to cover a larger envelope of Knudsen number. This work will present these exciting equations and the results obtained by solving them for various benchmark problems.

    

Effect of Swirler Configuration on Combustion of Steam Diluted Lean Premixed Methane-Hydrogen Mixtures in a Generic Dual Swirl Burner

Swirl-stabilized combustion of steam diluted hydrogen-enriched methane air mixture help in mitigating flashback and achieving low NOx and carbon emissions in lean premixed gas turbine combustors. Steam dilution reduces the flame temperature and speed associated with hydrogen addition, alters the concentration of active chemical species that forms NOx and allows for operating on high hydrogen content. The effect of hydrogen addition and steam dilution on flow fields, flame stability and NOx emissions can vary with swirl configurations. In this work, we numerically investigate the effects of swirl configurations (co- and counter-rotating) on flame stability and emissions of a dual swirl burner operating on premixed methane-hydrogen-air mixtures. Reynolds Averaged Navier Stokes (RANS) simulation approach employing Reynolds stress turbulence model and flamelet generated manifold combustion model is used to predict the flow within the burner and inside the combustor. Results based on characterization of axial and tangential velocity fields, temperature distribution, turbulent kinetic energy, NOx and carbon emissions are presented. Further, the flame shape is characterized by distribution of OH* and CH* radicals.   

Three-Dimensional Investigation of Particle Agglomeration using PIV

Proppant placement and settling characteristics significantly impact fracture conductivity. One key measure of reservoir permeability lies in proppant concentration at fractures. Despite its importance, minimal attention has been given to understanding the complex proppant agglomeration mechanisms, especially from a three-dimensional aspect. While two-dimensional studies indicate an inverse correlation between fluid viscosity and agglomeration, agglomerate forming mechanisms in a settling slurry are still ambiguous. The present study utilizes a combination of high-speed imaging and particle image velocimetry (PIV) to investigate the fundamental proppant agglomeration mechanisms in a three-dimensional domain. Multiple mesh size proppants are released in a rectangular transparent cell setup filled with varying solution viscosities. Addtionally, this study aims to identify an optimal proppant mesh size mixing ratio and fluid viscosity to facilitate hindered settling. Two high-speed cameras are set up for a full two-plane proppant PIV analysis and accurate agglomeration quantification. Preliminary results indicate unique agglomeration patterns associated with each solution viscosity and the existence of an optimal viscosity for reduced settling velocity with minimal agglomeration.

    

Numerical investigation of particle aggregate steering with magnetic resonance navigation for targeted embolization

Magnetic resonance navigation (MRN) has become a significant method to steer magnetized particles in medical applications by using MRI scanners. Dedicated MRN sequences have been successfully developed to track and steer such particle aggregates for selective chemoembolization of liver tumors. Particle aggregation is however a complex process and a comprehensive study of its shape and motion is needed to avoid damaging healthy tissues. Previous computational studies have considered non-aggregating particles or aggregates in 2D simplified settings. In this study, we investigate the forces acting on such micro-particle aggregation including drag, gravity, pressure gradient, virtual mass, magnetic gradient, dipole and contact forces. The code is developed in the OpenFOAM framework. Primary results are compared with bifurcation in vitro experimental data. Different sizes of clusters, directions of the magnetic field and gradient can be investigated to optimize aggregation form. This research represents an initial step towards complex in-vitro phantoms and in-vivo models.

    

Numerical Simulations of the Molecular Behavior and Entropy of Non-Ideal Fluids

A numerical model is built, simulating the principles of kinetic gas theory, to predict pressures of molecules in a spherical pressure vessel; the model tracks a single particle and multiplies the force on the spherical walls by a mole of molecules to predict the net pressure. An intermolecular attractive force is added for high-density simulations, to replicate a real fluid; the force is chosen to ensure the fluid matches the Peng-Robinson equation of state as it is compressed to a near supercritical density. This is completed for both the monatomic noble gas Argon, as well as the triatomic molecule carbon dioxide, with a significant Pitzer acentric factor. The standard deviations of the molecule position and velocity with respect to temperature and density is studied to define the entropy. A parametric study of a Stirling cycle heat engine utilizing near-supercritical densities is modeled, to study how the temperature dependence of the attractive inter-molecular Van der Waals forces can affect the net total entropy change to the surrounding environment.   

An Experimental Facility to Generate Homogeneous Isotropic Turbulence

We designed an experimental apparatus that generates homogeneous isotropic turbulence (HIT) to be used in fundamental studies to isolate the role of turbulence in complex flows and quantify its role in processes such as ice melting, flocculation, diffusivity and mixing. Inspired by the work of Bounoua et al. (2018), twenty synthetic jets are placed symmetrically around the tank exterior. We control the jets via the “random” forcing algorithm of Variano and Cowen (2008) to direct momentum to the center of the tank where a core of symmetric HIT absent mean flow is created. 

Particle image velocimetry (PIV) is used to measure turbulent flows within the experimental tank. Statistical analysis of PIV data provides mean and turbulent velocities, turbulent kinetic energy, spectra, integral scales, dissipation, Kolmogorov length and time scales, and the Taylor scale Reynolds number. The new facility required testing of a variety of parameter combinations of the jet forcing algorithm and pump outlet conditions including jet voltage, on time, and percentage of active jets to create an ideal turbulent environment. Initial measurements were completed to determine the length of an individual jet, varying voltage and on-time to inform HIT generation using the jet array.   

Anisotropic resistance effects on particle dispersion in turbulence

Particle-laden turbulent flows are important in both natural and industrial contexts. The particles in many of these processes, such as the formation of ice crystals in clouds or the paper-making process, are anisotropic, with directionally-dependent drag coefficients. Generally, anisotropic particles are free to rotate as they are advected by the carrier fluid. However, external forcing from gravitational and magnetic fields, the larger scale flow, and active behavior can restrict the particles’ orientation, fixing their anisotropic resistance with respect to the reference frame. The dynamics and statistics of symmetric particles in isotropic turbulence are well-studied, but the effect of anisotropic forcing on the transport and behavior of asymmetric particles is less well-understood. We studied these dynamics by conducting Lagrangian particle tracking in simulated isotropic turbulence from the Johns Hopkins Turbulence Database. Anisotropy was introduced by either directly scaling tracer-particle velocity or applying directionally-dependent Stokes drag and density parameters to particles simulated with the Maxey-Riley equations. We examine how increasing a particle’s resistance to motion in one direction in isotropic turbulence impacts the transport and dispersion statistics in all three directions.

    

Lagrangian Particle Statistics in a Homogeneous Swarm of Air Bubbles Rising in a Vertical Channel

Multiphase turbulent flows involving mixtures of water, oil, and gas are ubiquitous in nature and in industrial processes.  In systems utilizing multiphase flows for rapid mixing, it is necessary to predict the dilution and dispersion of passive tracers to assess the performance of the system. Using a set of direct numerical simulations (DNS) of a homogeneous air bubble swarm rising in a vertical channel, we attempted to test the validity of (1) G.I. Taylor’s 1921 turbulent diffusion law, and (2) particle-pair statistics as pertained to Richardson’s 1926 4/3th dispersion law in bubbly flows. The air void fraction in the simulations was fixed at 0.5% and we considered two scenarios – a monodispersed bubble swarm with 2mm-diameter bubbles and a bi-dispersed swarm with 2mm- and 4mm-diameter bubbles. We will compare our results to those from homogeneous isotropic turbulence (HIT), highlight the differences observed, and propose ideas to improve on existing turbulent diffusion theories.     

    

"Turbulent flow control through super-hydrophobic surfaces"

Shear flows, abundant in nature, can take the form of ocean, mantle, atmospheric convection, or other geophysical flows. Large-scale fluid structures, or secondary flows, can dominate the flows within these shear flows and cause changes to their surroundings; such a change can be found, for example, in atmospheric convections that bring rain, whereby secondary flows can facilitate pollutant transport which accelerates climactic events such as ozone depletion or global warming. The Taylor-Couette (TC) flow is a canonical model that scientists can use to study such secondary structures. The TC flow is described as a shear flow between two independently rotating co-axial cylinders. When secondary flows in the TC flow are pinned, they are called Taylor rolls; Taylor rolls can drastically affect flow behavior such as, for example, by hindering homogenous mixing even at high Reynolds numbers. We study the possibility of influencing these secondary structures, by interfering with the Reynolds stress that generates these Taylor rolls, in order to enhance mixing homogeneity. We achieve this by modifying the surface of the inner cylinder, using alternating super-hydrophobic (SHP) and no-slip surfaces in spanwise patterns on its surface at Rei = 104 to 2 × 104. We compare experiments of these inner cylinder surface modifications that affect the Taylor rolls and cause drag reduction.   

A simple model for the bottleneck effect in homogeneous isotropic turbulence based on Kolmogorov's hypotheses

We propose a model for the bottleneck effect, as a bump in the compensated energy spectrum, based on Kolmogorov's hypotheses (of 1941 and 1962). The model of the longitudinal structure function consists of two quadratic functions representing large- and small-scale motions. All the model parameters are either derived from asymptotic behaviors of the structure function or are universal constants fitted from direct numerical simulation (DNS) and experimental data. The model has a simple form without empirical parameters. From the model, the height of the spectral bump in the compensated spectrum has a power law R_λ^-0.0426, and the bump location scaled by the Kolmogorov scale is 0.153, which generally agree with various DNS results at moderate and large Reynolds numbers R_λ. Moreover, we derive that the incorporation of the intermittency exponent into the model leads to the decaying power law of the bump height with R_λ.   

Experimental study on the flow distribution and influence of the tidal stream generator to develop an underwater robot for maintenance

In this study, the flow speed distribution generated between two tidal stream power generators was visualized. A vortex with regularity was generated in the perpendicular direction of the blade rotation direction. It was observed that this could affect up to 10 to 12 times of blade diameter. On the other hand, in the case of the support structure and foundation, the vortex strength was lowered by more than 60%, and it was confirmed that both units had the same effect. Remotely Operated Vehicle(ROV) operating in a low vortex environment can be sufficiently controlled using the currently developed thrust-weight ratio. In the future, this study's results will be used to develop underwater robots for maintenance.   

Wind tunnel study of odor plume structure in the wake of a commercial odor-delivery device

We conduct laboratory experiments to quantify the spatiotemporal dynamics of odor plumes forming in the wake of a Training Aid Delivery Device (TADD). The TADD is a cylindrical glass jar topped with an odor-permeable membrane created for the US Army as an aid for training military working dogs for scent detection and localization. We use a photo-ionization detector (PID) instrument that is sensitive to vapors from volatile organic compounds (VOCs) to map negatively buoyant odor fields emanating from a TADD containing liquid acetone. The TADD is placed on the floor of a low-speed wind tunnel and an automated traverse translates the PID to measure three-dimensional odor concentration fields downwind of the TADD. Results show that the odor exiting the TADD membrane interacts with the horseshoe vortex structure formed by the interface between the bottom boundary layer and the TADD. This process creates an odor plume that has bimodal peaks away from the centerline and lower concentrations directly downwind. We investigate the effect of mean crossflow velocity in the tunnel and compare the results with numerical simulations presented in a companion poster.   

Numerical Simulation of Odor Plume Structure in the Wake of a Commercial Odor-Delivery Device

We conduct numerical simulations to quantify the spatiotemporal dynamics of odor plumes forming in the wake of a Training Aid Delivery Device (TADD). The TADD is a cylindrical glass jar topped with an odor-permeable membrane created for the US Army as an aid for training military working dogs for scent detection and localization. Simulations consist of a coupled solution of the Boussinesq form of the Navier-Stokes equations and the advection-diffusion equation. We simulate odors of varying specific gravities (SG) emanating from a TADD on the floor of a 3D rectangular domain with a cross flow, emulating related experiments performed in a low-speed wind tunnel and presented in a companion poster. We investigate odor plumes for a range of crossflow velocities and odorant SG. Results demonstrate that for SG=1, the odor plume is unimodal with peak concentrations on the centerline directly downstream. For SG>1, heavy plumes interact with the horseshoe vortex that forms around the TADD by creating a structured odor plume that has bimodal peaks away from the centerline and lower concentrations directly downwind.

Approved for public release. Distribution unlimited.   

Seismic Waves Propagation in Porous Media: A Pore-Scale Study

     Natural geomaterials and sediments contain a complex pore structure, with the skeletal portion of the material called the “matrix”. The soil as a porous medium itself can be assumed as a three-phase system that consists of a solid phase (soils), a gaseous phase (air), and a liquid phase (groundwater). The engineering properties of geomaterials are directly linked to the composition of matrix including the pore space structure and saturation. The transmission of seismic waves through these complex systems has been always an important research subject as the propagation of shear and compressional waves (S- and P-waves) can be used to explore the characterization of geomaterials in subsurface. 

     In this study, we study the seismic waves propagation in porous media through a pore-scale study. The goal of this work is to discover relationships between pore-scale characteristics media (distribution, heterogeneity and saturation) and seismic wave propagation in porous media. Random two-dimensional porous media are generated with different porosity (0 to 0.5), pore space heterogeneity (from uniform to highly heterogenous), and saturation (dried to fully water saturated). The models are imported into the Finite Element Method software package, COMSOL Multiphysics to simulate the propagation of seismic waves. A step pulse is transmitted on one boundary while the response is monitored via a probe on the other opposite side of the model geometry. The waves are propagated in both longitudinal, and horizontal directions to simulate pressure and shear waves. The Impacts of pore-scale characteristics of soils on shear- and pressure-wave travel times are determined and finally, a signal analysis methodology is suggested to identify both S-, and P-waves travel times in a complex porous medium by comparing the wave behavior with a non-porous medium.   

Unitary Wave Theory

This paper presents a unitary wave theory for the Macro-Universe to Meso-Nano to Micro-Fundamental particle in the Unitary Space-Time (UST). Using the nanoparticle diffraction X-ray peak profile analysis, specially according to Bragg's law, nλ = 2d Sin(θ), where θ - the Bragg angle, n- the order of diffraction, λ- the X-ray wavelength and d is the spacing between planes of given Miller indices. we construct the unitary wave theory, which can apply for all matters such as the Universe and Fundamental Particle. The Complex Circle Method (CCM) or Maximum Complex Torus in the Generalized Complex Structures is a strong quantum topological tool.

The reduced generalized complex structure (RGCS) in the complex circle is nothing but the quotient complex structure obtained from holomorphic quotient. Let ν be nanoparticle frequency, we have the Bragg's law: n V/ν = 2d Sin(θ). From the Generalized Newton's Laws (Zhji-An Luan, CAP 2019 Congress), we have V=4 Cos²(θ) and then the frequency 

ν = n 4 Cos²(θ) /(2d Sin(θ)) = 2n Cos²(θ)/(d Sin(θ)) = 2d/n Cot²(θ) Sin(θ),

which is just a Quotient Complex Structure of the complex circle. If (M, I) is a complex manifold, then any reduction of J_I by an extended action will result in a generalized complex structure and

J_I (T^*M) = T^*M

and the Courant Algebroid:

D= TM + T^*M

For a fixed mass M = Sin(θ), we finally obtain an Universal Frequency (UF):

v =  2βd/n Cot²(θ).

Here we obtain a monodromy law of all Matter- Wave, which means that the photon, electron, phonon and graviton havd a homotopy topological quantum structure, a Extended Dirac Instanton form:

Ω= (2n ± √3) + (2n±√3)^-1 = (2n ± √5) - (2n ± √5)^-1 |_n=1 = 4 (Maximum light speed > 3....km/s).

By the GNL theory:

GMνλ = 1 = Id, then ν = 1/(GMλ), where M- mass, G- generalized Newton's Constant Variety. If fixing a Standard Wavelength λ₀ = √3, we have the unitary wave form of

v = Ψ/(GM) ,where Ψ = 1/√3 is the Fu - Xi Universe Constant. 

This study provides a unique relation of Wave- Matter.   

Experimentally determining the reflection coefficient for reflecting internal waves

Determining the energy flux of an internal wave from the experimentally measured velocity field was made possible by the work of Lee et al. (Lee et al., Phys. Fluids, 26, 2014). This method is used in our work to measure the amount of energy dissipated when internal waves reflect from sloping boundaries by computing the reflection coefficient: the ratio of the outgoing energy flux to the incoming energy flux through a surface near to the reflection region. We account for viscous decay so that we can quantify the losses in the reflected wave due only to boundary processes and harmonic generation. We compare our experimental results to numerical simulations of the Navier-Stokes equations in the Boussinesq limit where the energy flux is known from the pressure and velocity fields. There is good agreement between our experimental and numerical simulation data, and we find that there are high rates of energy dissipation during reflection process. We also find that there is a wave reflected back from the boundary towards the generation site when either the boundary is rough or the angle of the boundary is close to the angle of the internal wave beam.   

Role of sea lion head in swimming and jumping

Sea lions are notably skillful swimmers. They use their fore flippers to generate thrust in a unique form of drag-based propulsion known as “clapping” and as hydrofoil-like structures to generate lift. Sea lions can complete a 180 degree turn in less than one body length, and they are known to rapidly exit and enter the surface of the water in a behavior known as “porpoising.” During these behaviors (swimming, turning, and porpoising) the effects of the head of these animals is unknown. Using 3D models of a sea lion head mounted in a wind tunnel, we examine the hydrodynamic forces generated by the sea lion head at different angles of attack. Separately, we use a drop cage and high-speed camera to study a sea lion head’s entry behavior at varying angles. These findings will help us understand how the sea lion head contributes to its behaviors, and inform future work in sea-lion-inspired aquatic vehicles.   

A Transition of the Intensity of Swirling of Vorticity Lines with Coalescence of Vortical Regions in a Homogeneous Isotropic Turbulence

The present study analyses the transition of coalescence of vortical regions in the core regions of vortices and traces the topology of bundles of vorticity lines using the local axis geometry. A coalescing phenomenon of vortices has been observed in a homogeneous isotropic turbulence, where the velocity field composes a large swirling motion in the swirl plane in the coalesced vortex. However, swirlity that specifies the swirling of the flow in terms of the local topology shows that this swirling motion is composed of two vortices in the same swirl plane. In addition, the vorticity lines swirl in the plane, and  the intensity of swirling increases at the coalescence part. This result suggests that the swirling feature of vorticity vectors may become intense due to the coalescing phenomenon. The vorticity vectors show a swirling feature when the non-diagonal components of the vorticity gradient tensor in the plane have the opposite signs in the swirl plane. The distributions of the components in the coalesced vortex show the complex feature in comparison with the non-coalesced vortex. It shows that the coalesced vortex is associated with a complicated phenomenon in terms of the topology of the vorticity vectors.

    

Nanoparticle enrichment in a sessile droplet using secondary acoustic radiation force

Nanoparticles are widely utilzed in many applications including nanomedicine and drug delivery. Although the nanoparticle enrichment is fundemantal in nanoscience and nanotechnology, the conventional centrifugation and filtration methods are difficult to apply for the nanascale objects. Here, we propose an acoustic method to enrich the nanoscale objects using secondary acoustic radiation force. When MHz-frequency acoustic waves are applied to a mixture of micro- and nanoparticles are suspended in a sessile droplet, the incident waves are scattered off the microparticles; the nanoparticles are captured and enriched in between the microparticles by secondary acoustic radiation force. We experimentally demonstrated that the 700 nm polystyrene (PS) nanoparticles can be captured using the 7 μm PS microparticles suspended in a 3 μl water sessile droplet subject to a 72.5 MHz acoustic field within a few tens of seconds. We expect that the proposed method can be applied to droplet-based point-of-care testing using nanoscale biomolecules.   

CFD Visualization using Augmented Reality

We leverage the immersive and interactive nature of Visualization techniques such as Augmented Reality (AR)) have been around and are incorporated in various fields to utilize its interactive and immersive nature for optimization, design purposes, and understanding complex phenomena. With the field of to make visualization of fluid flows more intuitive and engaging for undergraduate studentsFluid Mechanics being no different, AR has gradually extended its roots and captivated the attention of several researchers who want to explore this idea further. For this research project, we perform a large-scale simulation of flow around a quadcopter. While the CFD simulation provides a wealth of data, visualization using traditional methods on two-dimensional displays does not provide a thorough understanding of the ongoing phenomenon as many intricate three-dimensional features may be hidden. To remedy this problem, we use Blender and Unity to transform vorticity and pressure iso-surfaces into AR objects that can be visualized using a consumer smartphone. We show how the AR visualization makes the task of inspecting flow features intuitive and straightforward. This engaging technique can be utilized in outreach programs to educate students about Fluid Mechanics.

 

Funding Acknowledgement:

This work is supported by the National Science Foundation under grants CBET-FD-2028617 and EEC-GOALI 2216969.

    

Airfoil Passive Flow Control Strategies: Leading-Edge versus Surface Modifications

The aerodynamic and flow field characteristics of a NACA 0010 airfoil with and without modifications was investigated in flow with freestream turbulence intensity of 4%. The airfoil was modified with two different passive flow control strategies: leading-edge tubercles over the spanwise of the airfoil and surface-mounted rotating cylinders. The unmodified version of the airfoil was used as a baseline for comparison over a range of Reynolds numbers from 197,200 to 250,000 and various angles of attack ranging 0° to 24°. The experiments were carried out using two 6-axis force/torque transducers and Particle Image Velocimetry (PIV) in the low-speed wind tunnel facility at the University of Texas – Rio Grande Valley. The flow field data obtained using PIV showed large amount of turbulent mixing at the leading edge of the tubercle airfoil configuration. The PIV data revealed the presence of a recirculation bubble when the rotating cylinders on the surface of the airfoil were activated, giving rise to a faster adverse pressure gradient for the reattachment of the separated boundary layer. The recirculation greatly increased lift performance up to 50% within the pre-stall regime when compared to the unmodified airfoil and the tubercle airfoil.   

The Fluid Mechanics of Bacterial Motility

Many bacteria have evolved a motility system that consists of a helical appendage attached to the cell body by a rotary molecular motor. The motor is driven by protons flowing from the outside to the inside of the cell. The resulting rotation of the helix propels the bacterium through its low Reynolds number fluid environment. In this work, we determine the energy cost of bacterial motility by inputting experimentally measured bacterial motion into computational fluid dynamics simulations. The computational method is precisely calibrated using macroscopic fluid dynamics experiments to ensure the calculated energy values are accurate. We compute the energy cost of motility over a range of body and helix geometries, and we find that the geometries realized by living bacteria are near the minimum value of this measure.