It has previously been shown that by increasing the Reynolds number across a channel by spatially varying the viscosity does not cause an immediate change in the size of turbulent structures and a delay is in fact observed in both wall shear and friction Reynolds number (Coppo Leite, V, & Merzari, E., Proceedings of the ASME 2020 FEDSM, p. V003T05A019). Furthermore, it is also shown that depending on the length in which the flow condition changes, turbulence bursts are observed in the turbulence field. For the present work we propose a new version of the standard Reynolds Averaged Navier Stokes (RANS) k – τ model that includes some modifications in the production term in order to account for these effects. The new proposed model may be useful for many engineering applications as turbulent flows featuring temperature gradients and high heat transfer rates are often seen in heat exchangers, combustion chambers and nuclear reactors. In these applications, thermal and viscous properties of the working fluid are important design parameters that depend on temperature; hence it is likely to observe strong gradients on these scalars’ fields. To accomplish our goal, the modifications for the k – τ model are implemented and tested for a channel flow with spatial varying viscosity in the streamwise direction. The numerical simulations are performed using Nek5000, a spectral-element code developed at Argonne National Laboratory (ANL). Finally, the results considering a turbulence channel using the proposed model are compared against data obtained using Direct Numerical Simulations from the earlier work.

Predicting aerodynamic forces on bluff bodies remains to be a challenging task due to the unpredictable flow behavior, specifically at higher Reynolds numbers. Experimental approaches to determine aerodynamic coefficients could be costly and time consuming. In the meantime, use of numerical techniques could also require a considerable computational cost and time depending on complexity of the flow behavior. The research focusses on developing an effective deep learning technique to predict aerodynamic force coefficients acting on elliptical bluff bodies for a given aspect ratio and given flow condition. Collecting data for drag and lift coefficients of several aspect ratios for flow conditions starting from onset of vortex shredding to verge of subcritical region is conducted by an accurate full order model. The specified region will provide a transient flow behavior and thus lift coefficient will be represented in terms of root mean square value and drag coefficient in terms of a mean value. With variations in flow behavior and vortex shredding frequencies, it requires to select an appropriate turbulence model, optimum discretization of fluid domain and time step to obtain an accurate result. Flow simulations are conducted primarily using Unsteady Reynolds Averaged Navier-Stokes Equations (URANS) model and Detached Eddy Simulations (DES) model. Effectiveness in using different turbulence models for specified flow regimes are also explored in comparison to available experimental results. At lower Reynolds numbers, aerodynamic force coefficients for a specified body will only depend on Reynolds number. But after a certain specific Reynolds number, aerodynamic forces are dependent on the Mach number in addition to Reynolds number. Therefore, for higher Reynolds numbers, aerodynamic force coefficients are recorded for multiple Mach numbers with same Reynolds number and will be fed to the neural network. With the development of the machine learning and neural network modelling, many of the fields have nourished and created effective and efficient technologies to ease complex functions and activities. Our goal is to ease the complexity in the computational fluid dynamic field with a deep neural network tool created to predict drag and lift coefficient of elliptical bluff bodies for a given aspect ratio with an acceptable accuracy level. Researchers have developed deep neural network tools to predict various flow conditions and have succeeded with sufficient accuracy and a satisfying reduction of computational cost. In our proposed deep learning neural network, we have chosen to model the network with inputs as the geometry setup and the flow conditions with validated drag and lift coefficients. The model will extract the necessary flow features into filters with the convolution operation performed on the inputs. Our main directive is to create a deep learned neural network tool to predict the target values within an acceptable range of accuracy while minimizing the computation cost.

Prediction of the drag coefficient is required in gas-particle multiphase flow modeling and simulation. Experimental data and correlations on the fixed-bed system of spherical particles with high volume fractions for various possible arrangements are available in the literature. However, the effect of volume fraction on the drag coefficient of non-spherical particles is not well studied. In solving the momentum equation, the volume fraction plays a vital role in determining the flow resistances. In this paper, we study the impact of volume fraction in the range of 0.069 to 0.65 on the drag coefficient using the computational fluid dynamics (CFD) simulation of air for Reynold number in the range of 10 to 10000 using particle resolved direct numerical solution (PR-DNS). Regular non-spherical particles such as a cube, tetrahedron, and spheroids are used in this study since their single particle’s drag coefficient data are available in the literature for comparison. For this work, the simulations are carried out in the Ansys Fluent using polyhedral mesh, which consumes significantly less computational time and power. The study showed the sphericity and volume fraction have significant impact on the bed pressure drop and average drag coefficient of the particles in the bed especially in high Reynolds number regime. The bed of the spheroid experiences the lowest drag being the most streamlined particle, and the particles with the edges result in a large drag coefficient due to flow separation at the discontinuity. The vector plots verify this behavior where large wake regions are observed behind the tetrahedron particle.

Understanding the transition from laminar to turbulent flow – Boundary-Layer Transition (BLT), we can design better state-of-the-art vehicles for defense and space applications, which can mitigate the limitations in current high-speed temperature conditions. BLT is a subject of fluid flow disturbances created by geometric parameters and flow conditions, such as surface roughness, increased velocity, and high-pressure fluctuations, to name a few. These disturbances lead to the development of turbulent spots and differential heating. Historically, the Reynolds number has been used to predict whether a system will develop turbulent flow. However, it has been known for decades that it is not always reliable and cannot indicate where the BLT will occur: some experiments present scenarios where the flow is laminar at a high Reynolds number and vice versa. We can predict the BLT from performing physical experiments, but they are expensive and physical configurations are limited. Despite many community efforts and successes, no general computational solution to simulate different flows and vehicle types that fully incorporate BLT exists. Many are a considerable number of parameters that affect BLT. Therefore, we use Causal Inference to predict BLT by cause-and-effect analysis on multivariate data obtained from BLT studies. Data generated using high-fidelity Computational Fluid Dynamics (CFD) with resolved Large-Eddy Simulations (LES) scales, will be analyzed for turbulence intensity by decomposing velocity in mean and fluctuations. In this paper, we will be discussing approaches on how we predict BLT scenarios using cause and effect relationships driven by causal inference analysis.

In 2011, John Dabiri proposed the use of counter-rotating vertical-axis wind turbines to achieve enhanced power output per unit area of a wind farm. Since then, various studies in the wind energy and marine hydrokinetic (MHK) literature have been dedicated to pairs of vertical axis turbines in both co-rotating and counter-rotating configurations, in terms of their power production, wake characterization, and optimal array design. Previous experimental works suggest an enhancement of up to 27.9% in the system power coefficient of pair configurations compared to a single turbine. Additionally, previous numerical studies have indicated that the increased power output is correlated with higher torque on the turbine blades which correspondingly produces a stronger leading edge vortex. This paper presents an extended investigation into a pair of laboratory scaled cross-flow hydrokinetic turbines in counter-rotating configurations. Experiments were conducted to observe, compare, and discuss the leading edge vortex shedding from the turbine blades during their positive torque phase. The turbines operated in a small water flume at the diameter-based Reynolds number of 22,000 with a 0.316 m/s freestream velocity and 4% turbulent intensity. Using a monoscopic particle image velocimetry setup, multiple realizations of the water flow around each blade at their positive torque phase were recorded and phase-averaged. Results show consistent leading vortex shedding at these turbine angles while a correlation between the turbine power performance and the vortex size and strength was observed.

Transverse jet from elevated source is found in various environmental and industrial field which include smoke exhausting from stack into atmosphere and sewage water disposal in deep-ocean. The experiment is carried out in water tunnel using flow visualization and Laser Doppler Velocimetry. Analysis has been performed for axisymmetric round jet of aspect ratio of 9.0 with the velocity ratio varying up to 2.5 at Reynolds number (based on free stream crossflow velocity and jet external diameter) of 1000. Result shows the formation of different jet shear layer vortices with varying velocity ratio are: (i) clockwise-downwash vortices (velocity ratio less than 0.3), (ii) delayed-regular-clockwise vortices (between 0.3 to 0.7), (iii) regular-clockwise vortices (between 0.7 to 1.4) at the lee side of the jet shear layer, (iv) irregular-anticlockwise vortex at the upstream along with clockwise vortices at the lee side of the jet shear layer that together forms mushroom vortices (between 1.4 to 1.9) and (v) regular-mushroom vortices (above 1.9). The other vortices found are stack-end vortex (less than 0.9) in the wake near free end, upright vortices (above 0.9) in jet-wake and hairpin vortices (between 0.3 to 0.6) in downstream which is the stretched part of evolving shear layer.

For a tandem airfoil configuration, an airfoil is placed in the wake of an upstream airfoil. This interaction affects the aerodynamic forces of the airfoils, especially the downstream one. In the present study a tandem configuration consists of an upstream pitching airfoil and a downstream stationary airfoil is investigated. This study aims to investigate the role of reduced frequency and pitch amplitude of the upstream airfoil’s motion on lift and drag coefficients of two airfoils. These two parameters play an important role in the formation of vortices. The investigation is done for Selig-Donovan 7003 (SD7003) airfoils at low Reynolds number of 30,000 using a computational fluid dynamics. Incompressible URANS equations were employed for solving the flow field. It was found that for a fixed reduced frequency of 0.5 thrust is produced on the hindfoil for a part of cycle for different pitch amplitudes from light to deep stall while for a fixed pitch amplitude at different reduced frequencies high level of thrust or drag can be produced. The reason is related to the type and intensity of vortex-blade interaction.

3-D numerical simulations are performed to investigate liquid sodium (Na) flow and the heat transfer within miniature heat sinks with different geometries and hydraulic diameters of less than 5 mm. Two different straight small-scale heat sinks with rectangular and triangular cross-sections are studied in the laminar flow with the Reynolds number up to 1900. The local and average Nusselt numbers are obtained and compared against eachother. At the same surface area to volume ratio, rectangular minichannel heat sink leads to almost 280% higher convective heat transfer rate in comparison with triangular heat sink. It is observed that the difference between thermal efficiencies of rectangular and triangular minichannel heat sinks was independent of flow Reynolds number.

Buoyant shear layers are encountered in many engineering and environmental applications, and have been studied by researchers in the context of experiments and modeling for decades. Often, these flows have high Reynolds and Richardson numbers, and this leads to significant/intractable space-time resolution requirements for DNS or LES modeling. On the other hand, many of the important physical mechanisms in these systems, such as stress anisotropy, wake stabilization, and regime transition, inherently render eddy viscosity-based RANS modeling inappropriate. Accordingly, we pursue second-moment closure (SMC), i.e., full Reynolds stress/flux/variance modeling, for moderate Reynolds number non-stratified, and stratified shear layers for which DNS is possible. A range of sub-model complexity is pursued for the diffusion of stresses, density fluxes and variance, pressure strain and scrambling, and dissipation. These sub-models are evaluated in terms of how well they are represented by DNS in comparison to the exact Reynolds averaged terms, and how well they impact the accuracy of the full RANS closure. For the non-stratified case, the SMC model predicts the shear layer growth rate and Reynolds shear stress profiles accurately. Stress anisotropy and budgets are captured only qualitatively. Comparing DNS of exact and modeled terms, inconsistencies in model performance and assumptions are observed, including inaccurate prediction of individual statistics, non-negligible pressure diffusion, and dissipation anisotropy. For the stratified case, shear layer and gradient Richardson number growth rates, and stress, flux and variance decay rates, are captured with less accuracy than corresponding flow parameters in the non-stratified case. These studies lead to several recommendations for model improvement.

The interactions between the separated shear layer and the near wake region of rectangular cylinders of varying streamwise extents in a uniform flow are investigated using time resolved particle image velocimetry. The streamwise aspect ratios (AR) tested were 1 and 5, and the Reynolds number based on the oncoming flow velocity and cylinder height is 16200. The effects of varying AR on the mean flow, turbulent kinetic energy and Reynolds shear stresses are studied. Furthermore, the unsteady characteristics of the separation bubbles are examined in terms of frequency spectra analysis. The mean flow topology shows flow separation at the leading edge is not affected by the streamwise aspect ratios. However, the primary, secondary and wake vortexes show significant differences. Mean flow reattaches over the cylinder at 4.30 cylinder heights in the AR5 case while there is no mean reattachment in the AR1 case. The magnitudes of turbulent kinetic energy and Reynolds shear stress in the wake region are an order of magnitude higher in AR1 compared to AR5. Depending on the streamwise location, the vortex shedding motions in the near wake region reflect the dominant and second harmonic of the shear layer shedding frequency measured near the leading edge.

In this paper, we evaluate the hypothesis that bump arrays can be used to separate particles from turbulent flows entering the array. Microfluidic bump arrays are known for separating particles by size from laminar inlet flows. However, turbulent inlet flows have not been explored but become important as microfluidic bump arrays are scaled up to mesofluidic bump arrays. We find experimentally that particle separation is indeed effective at higher Reynolds numbers. These experimental findings portend industrial scale particle separation due to the higher flow rates they facilitate.

Turbulent heated and buoyant plumes have important applications in the atmosphere such as wildland fire plumes, volcanic plumes, and chemical plumes. The purpose of the study is to analyze the turbulence structures, and to understand the stages of the development of the starting turbulent plumes. For this purpose, data generated from an in-house Weather Research Forecast model coupled with Large-eddy simulation (WRF-bLES) with two-way feedback between the buoyant plume and the atmosphere developed has been used. The release of both dense gases (Co2, So2) and, buoyant gases (He, NH3, heated air) from a circular source at the bottom of the domain have been investigated. The simulations of the axisymmetric plume were performed at a high Reynolds number of 10 8 . Vortex Identification methods were used to extract the Coherent structures and the large-scale features of the flow. The results have demonstrated that both the dense and the buoyant heated plumes with different initial characters exhibited universal characteristics and the development of the starting plumes occurred in four characteristic stages: Stage 1 is the plume acceleration stage, followed by stage 2 which corresponds to the formation of the head of the plume which grows spatially. Stage 3 is when the plume head is fully formed and the flow transitions to quasi-steady-state behavior. The final stage is the fully developed plume. The identification of the four-stage development of the plume in the neutral environment is the first step in studying the turbulent heated and buoyant plumes development in order to characterize realistic plumes and to quantify the extent of mixing at each of these stages. This work has important contributions to fundamental fluid dynamics of buoyant plumes with implications on forecasting the plume trajectory of smoke, wildland fire, and volcanic plumes.

3D laminar flow simulations were conducted using OpenFOAM to resolve the temperature, concentration, velocity, and pressure field for two hollow fiber vacuum membrane distillation configurations with feed solution flowing either inside or outside a single hollow fiber. The fiber has a circular cross-section, a fixed length of 120mm, and an inner diameter of 2 mm. The wall thickness was varied from 150 to 450μm, and the pore diameter was varied from 0.1 to 0.3μm based on commercial fibers. The feed solution is an aqueous solution of water and NaCl. The feed flow was simulated at a Reynolds number of 500 and vacuum pressure of 5,000 Pascals. It was found that there was a 75% increase in flux, from 9.58 to 41.41 kg/m 2 h, between the worst and the best case in membrane properties. Increasing the pore diameter or wall thickness while the other value was fixed resulted in a 45–57% flux increase depending on the fixed value. The module with the feed solution flowing outside a hollow fiber yields 24% higher flux than the module with the feed solution flowing inside the hollow fiber at the same conditions.

Computational fluid dynamics simulations were conducted to evaluate the effect of a wavy channel in a two-dimensional vacuum membrane distillation module. The curvature was induced using a sinusoidal profile along the membrane and bottom wall. Contour plots and line profiles provide a detailed view of the flow structure and the effect of the proposed configuration on the flux performance. Module averaged temperatures, concentrations, and flux values were calculated for two selected Reynolds numbers. Results indicate that at low Reynolds numbers, the wiggly module performs worse than the flat sheet module. Due to the channel’s curvature changing the direction of the bulk flow and the absence of secondary flows to promote mixing, the thermal boundary layer along the membrane surface can be more intense versus a flat sheet membrane, causing more intense temperature polarization and reduced flux performance. At Reynolds number 500, there was a 5% decrease in the flux for the curved versus flat case. However, in some curved channel areas, the local performance was superior to the flat channel. Increasing the Reynolds number could aid the performance of the wiggly channel immensely.

We perform RANS-VOF simulation of density-stratified, fully developed air-water flow in a 2D channel. The flow is completely specified by the (common) driving pressure gradient down the channel and by the fill factor (relative height of the heavier phase to the total height of the channel). Varying the pressure gradient and fill factor results in different flow combinations: namely, laminar air/laminar water, turbulent air/laminar water, turbulent air/turbulent water, laminar air/turbulent water. The focus of the study is the near-interface interaction when either or both phases are turbulent. The RANS-VOF equations are solved on a 2D channel with periodic inlet/outlet boundary conditions. For a fixed fill factor, the Reynolds numbers of each phase varies monotonically with driving pressure gradient. However, at fixed pressure gradient, the fluid Reynolds numbers are non-monotonic at high fill factor. This Reynolds number phase diagram is important in understanding the laminar and turbulent regimes of each phase. The mean velocity is axial (down the channel) and exhibits a dip below the free surface whenever one of the phases is turbulent. As expected, the diagonal Reynolds stresses <u’u’>, <v’v’>, are strongly coupled between the phases, however coupling in <u’v’> is less pronounced.

The present paper presents some experimental and numerical modeling of the Newtonian and viscoelastic flows in bifurcated configurations of pipes, for stationary and non-permanent regimes. The main purpose of this study is to select an optimal design of the pipes ramifications, for minimizing the local pressure losses and to improve the efficiency of fluid transportation. The method is based on the transformation of the flow field obtained with particle image visualization technique, for different Reynolds numbers. Based on experimental results will be established optimum geometry of the ramifications. The results will be used as the boundary conditions for numerical modeling. The experimental measurements were performed into a closed circuit of pipes, with different diameters, consisting of a centrifugal pump supplied by a tank, sensors for estimating the pressure losses and devices for measuring the flow rate. It is detailed presented in a dedicated paragraph. The main pipe is connected to a transparent bifurcation with branches at different angles from the main pipe axis. The measurements illustrate that the flow has different aspects, depending on the bifurcation’s angle. The numerical simulations are performed with Fluent CFD based on the volume numerical method, to obtain the Navier-Stokes solutions for the Newtonian model in the laminar or turbulent flow conditions. A pre-processor has been used to create the geometry of the bifurcation and to generate the mesh. The 3D-flow domain contains 944390 volumes, tetrahedral hybrid. It was obtained the numerical solutions of the fluid flow in branching pipes for the Reynolds numbers from 1000 up to 40000. The governing equations were assumed from the k-ε model for turbulence flow, the equation of continuity, equation of fluid motion, and the transport equation. Finally, some conclusions and references are presented.

At several natural gas compressor stations control valves have experienced low-frequency vibrations when placed directly downstream of pipe bends. It is know that at pipe bends a pair of counter-rotating vortexes form and extend several diameters downstream. These vortexes are not symmetric and instead oscillate back and forth at a low-frequency phenomenon referred to as swirl-switching. It is suspected that these upstream flow oscillations enter the valve and are the cause of the valve vibrations. Large-eddy simulations have been performed on pipe bends with two different bend radius at a Reynolds number of 80,000. The simulations showed that the frequency of the vortexes decreased as the bend radius increased. When non-dimensionalized the frequency obtained from the simulation is comparable to the field measurement from the compressor station. This consistency provides support to the argument that upstream flow-induced oscillations are the cause of the low-frequency valve vibrations. The velocity fluctuations due to swirl-switching dissipate downstream of the bend and by placing several diameter lengths of straight pipe between a control valve and a pipe bend, the issue of low-frequency vibrations induced at the valve could be avoided. The objective of this study is to provide evidence that low-frequency valve vibrations can be linked to upstream pipe geometry and put forward a plausible explanation of why.

Vertical takeoff and landing vehicle platforms with many small rotors are becoming increasingly pertinent for small Unmanned Aerial Vehicles (UAVs) as well as distributed electric propulsion for larger vehicles. These rotors operate at low Reynolds number unlike large rotors for which the existing prediction methods were developed. Operating at very low Reynolds number essentially means that viscous effects are more dominant; and their spatial spread is significant with respect to the rotor dimensions. This impacts the nature of inter-rotor aerodynamic interactions which become more difficult to predict and characterize. In the present research, two nominally identical commercial UAV rotors are studied for a range of separations in hover and forward flight, both experimentally and computationally, in parallel with ongoing vehicle flight tests with 4 and 8 rotors. Bi-rotor tests in tandem in-plane configuration were performed in Georgia Tech’s 2.13m × 2.74m test section wind tunnel. Rotor simulations were done using the RotCFD Navier-Stokes solver. In hover, rotor performance is sensitive to the distance between rotors at low rotation speeds, indicating the presence of greater inter-rotor interactions at low Reynolds number. In forward flight, the performance of the downstream rotor gets negatively affected by the upstream rotor wake.

In this paper, we use a CFD analysis of a simplified, 2D geometry to study the ability of mako shark denticles to mitigate flow separation. We represent the viscous sublayer below a turbulent boundary layer streak as a Couette flow. Incipient separation is simulated by balancing upper wall velocity and adverse pressure gradient to achieve zero net mass flow, and we add various denticle geometries to study their effects. Each modeled denticle protrudes at an angle from 15° to 85° and sublayer blockage ratio from 0.05 to 0.85. Through variation of fluid properties and boundary conditions, we show that the anti-flow-reversal abilities of a single, bristled shark denticle are independent of Reynolds number, and we investigate the effect of the denticle at cases other than zero net mass flux. Based on these results, we create a new relationship to predict separation inhibition. These conclusions are highly generalizable and represent previously undiscovered universal behavior.

Flow conduits made via additive manufacturing, commonly referred to as 3-D printing, are of increasing interest for a variety of industrial applications due to the ability to create unique and conformal flow paths that would not be possible with other fabrication techniques. Fused filament fabrication (FFF) is an additive manufacturing technique that is seeing new interest in the creation of internal flow channels with its ability to print high-temperature polymers and soluble supports. Printing parameter choices in the FFF printing process result in surfaces that can have significant profile differences that may significantly impact the flow characteristics within the conduits. In this study, two print parameters were experimentally studied for turbulent water flow through circular pipes created by fused filament fabrication out of acrylonitrile butadiene styrene (ABS). The print layer orientation relative to the flow was investigated for printing layers parallel, perpendicular, and at 45 degrees from the flow axis. Layer thickness were varied from 0.254 mm to 0.330 mm and all channels were created using soluble support structures. Pressure drops were measured for fully developed flow through pipes with an inside diameter of 5 mm and Reynolds numbers up to 62,000. Results are presented in terms of relative pressure drops as well as the wall surface roughness that would lead to such impacts. These flow-determined grain surface roughnesses are then compared against measurements of print surface roughness.