The presence of complex vortical structures, unsteady wakes, separated shear layers, and streamline curvature pose considerable challenges for traditional linear Eddy-Viscosity (LEV) models. Since Non-Linear Eddy Viscosity Models (NEV) models contain additional strain-rate and vorticity relationships, they can provide a better description for flows with Reynolds stress anisotropy and can be considered to be suitable alternatives to traditional EVMs in some cases. In this study, performance of a Non-Linear Explicit Algebraic Reynolds Stress Model (NEARSM) to accurately resolve flow over a surface mounted cube and a 3D axisymmetric hill is evaluated against existing experimental and numerical studies. Numerical simulations were performed using the SST k-ω RANS model, SST k-ω -NEARSM, SST-Multiscale LES model, and two variants of the Dynamic Hybrid RANS-LES (DHRL) model that include the SST k-ω and the SST k-ω -NEARSM as the RANS models. Results indicate that the SST k-ω RANS model fails to accurately predict the flowfield in the separated wake region and although the SST-NEARSM and SST-Multiscale LES models provide an improved description of the flow, they suffer from incorrect RANS-LES transition caused by Modeled Stress Depletion (MSD) and sensitivity to changes in grid resolution. The SST-DHRL and the SST-NEARSM-DHRL variants provide the best agreement to experimental and numerical data.

In this paper, coupled CFD-DEM simulations of dense particle-laden jet flow are performed using CFDEM ® coupling interface that couples LAMMPS-based LIGGGHTS ® DEM engine with OpenFOAM CFD framework. Suspensions of mono-sized spherical glass particles with 80 microns diameter and a mass loading of 0.23 and 0.86 are considered. Three different CFD meshes are used with an average mesh resolution dimension of 3.06, 2.67, and 1.86 particle diameters and it is determined that mesh resolution does not change results for void fraction calculation (using the divided model) of the CFD-DEM equations. Samples of particle flux are taken at 0.1, 10, and 20 nozzle diameters along the axial direction of the jet region. The numerical results for particle flux are compared with a well cited experimental data found in literature. The CFD-DEM simulations in turbulent jet flow are found to be highly sensitive to initial particle velocity inputs but the experimental data provide sufficient information to produce comparable results.

Computational fluid dynamics (CFD) results are presented for synthetic turbulence generation by a proposed statistically targeted forcing (STF) method. The new method seeks to introduce a fluctuating velocity field with a distribution of first and second moments that match a user-specified target mean velocity and Reynolds stress tensor, by incorporating deterministic time-dependent forcing terms into the momentum equation for the resolved flow. The STF method is formulated to extend the applicability of previously documented methods and provide flexibility in regions where synthetic turbulence needs to be generated or damped, for use in engineering level large-eddy and hybrid large-eddy/Reynolds-averaged Navier-Stokes CFD simulations. The objective of this study is to evaluate the performance of the proposed STF method in LES simulations of isotropic and anisotropic homogeneous turbulent flow test cases. Results are interrogated and compared to target statistical velocity and turbulent stress distributions and evaluated in terms of energy spectra. Analysis of the influence of STF model parameters, mesh resolution, and LES subgrid stress model on the results is investigated. Results show that the new method can successfully reproduce desired statistical distributions in a homogeneous turbulent flow.

The Cahn-Hilliard phase field method for two-phase flow has gained particular attention due to its unique features including its flexibility for complex morphological and topological changes, the intrinsic property of conserving mass, and the natural approach to account for the surface tension. The essential idea of the method is to use a phase field function to describe the two-phase system, while a thin smooth transition layer (interfacial area) connects the two immiscible fluids, where the value of phase field function varies continuously. The application of phase field method to two-phase flows has become more widespread recently, but to the best of our knowledge, very little progress has been made for the method being applied to the two-phase flows with phase change. This includes evaporation, condensation and boiling, which plays an important role in enhanced heat transfer in power electronics, energy, and aerospace engineering. In previous work, in order to face the challenge of large density contrast and high Reynolds number of practical engineering problems, we developed a stabilized phase field method that can handle two-phase flow with density ratio over 1000, at high Reynolds number over 10,000, and applied the method to simulate slug initiation in a long circular pipe. In this paper, inspired by the boiling model widely used in the level-set method, we propose a new boiling model that assumes that boiling takes place in the whole interfacial layer. The method is then used to solve the non-solenoidal Navier-Stokes equations. The boiling model is validated by simulating a vapor bubble growing in super-heated liquid. For this case, the growth rate of the bubble has an analytical solution, and it is used as a benchmark case in volume of fluid (VOF) and level-set methods extensively. For three different refrigerants, namely water, R134a and HFE7100, our phase field method with the boiling model can obtain accurate simulation results. Moreover, the method and model are applied to predict the three-dimensional boiling heat transfer in a rectangular micro-channel that contains a water vapor bubble with various inlet super-heat conditions. We found that the predicted bubble shape is very similar to that visualized in existing experiment. From our simulation of boiling flow using the phase field method, We have found that the required mesh resolution for the phase field method is comparable with that of VOF and level-set methods.

We investigate electrokinetic behavior of colloidal particles in the vicinity of a solid-liquid interface. Colloidal liquids are expected to be used as thermal transport media for heat transfer applications such as nanofluids and phase change emulsions. They contain submicrometer-sized particles in liquid, and electrokinetic behavior of the solute particles should play an important role in the heat transfer between solid-liquid interfacing boundaries. However, experimental investigation of the behavior remains difficult due to the required spatial resolution beyond diffraction limit. We developed a measurement system based on laser Doppler principle using an interference of evanescent waves generated at total internal reflections of incident lasers at a solid wall. The system was developed for the measurement of velocities of colloidal particles at an interfacing boundary of colloidal liquid and a solid wall. The system has a unique advantage of a high spatial resolution in the direction perpendicular to the boundary due to the short penetration depth of an evanescent wave in the range of a few hundred nanometers. The principle and performance of the measurement system were investigated using a scanning probe in the measurement volume. We experimentally confirmed the validity of the measurement and characterized the uncertainty of velocity measurement. The system was further applied in a series of measurements of alumina particles dispersed in water in a square-shaped cell under induced electric fields. The measured velocities are proportional to the field strengths at different particle concentrations. The linear relationship is consistent with theoretical predictions, which demonstrates the feasibility of the system for the measurement of velocities of colloidal particles in the near wall region.

Decay of honeycomb-generated turbulence in a duct with a static transverse magnetic field is studied via high-resolution direct numerical simulations. The simulations follow the revealing experimental study [1], in particular the paradoxical observation of high-amplitude velocity fluctuations, which exist in the downstream portion of the flow when the strong transverse magnetic field is imposed in the entire duct including the honeycomb exit, but not in other configurations. It is shown that the fluctuations are caused by the large-scale quasi-two-dimensional structures forming in the flow at the initial stages of the decay and surviving the magnetic suppression. Statistical turbulence properties, such as the energy decay curves, two-point correlations and typical length scales are computed. The study demonstrates that turbulence decay in the presence of a magnetic field is a complex phenomenon critically depending on the state of the flow at the moment the field is introduced.

In this contribution, an assessment of compressible Reynolds Averaged Navier Stokes equations (RANS) and Large Eddy Simulation (LES) is presented using transonic organic vapor flow past a NACA4412 airfoil as a case study. The NACA4412 represents a canonical geometry, which, in case of air, has been well investigated numerically and experimentally. The results of the real gas simulations are compared with those of air simulations. For the real gas, the organic vapor Novec 649 ® is chosen as a representative fluid. The thermodynamic behavior of Novec 649 ® is modeled with the Peng-Robinson equation of state. Different inlet Mach numbers are applied, namely, a sub-critical, the critical, and a super-critical Mach number. It turns out, that the critical Mach number of the NACA4412 airfoil increases when Novec 649 ® is used as working fluid. Furthermore, it is shown that real gas flow simulations cause additional difficulties for the computational fluid dynamics (CFD) analysis. Although the speed of sound of Novec 649 ® is lower than the speed of sound of air, a finer grid resolution is required for the real gas simulations due to its high density. Based on an extensive simulation study, an assessment of different numerical modelling strategies and methods is given.

Ventricular assist devices (VAD) are designed to provide circulatory support to patients suffering from advanced-stage heart failure. While not pulsatile, the advantages of continuous axial flow VADs include a compact size and low mechanical failure rate. However, being compact, they operate at high speed, resulting in adverse effects, such as hemolysis caused by high flow shear and thrombosis formation in stagnant regions are common and threaten the successful use of the device. While state-of-the-art computational fluid dynamics (CFD) is widely used in designing these devices, detailed high-resolution experimental measurements of the flow within them are not readily available in the literature. Such experimental data is crucial for understanding the flow inside the VAD and its interaction with blood cells as well as for validating the CFD predictions. The present study investigates the flow inside a 1:1 exact replica of a VAD – ReliantHeart HeartAssist5 ® . This 12mm diameter device consists of an inlet guide vane (IGV), a rotor and a stator. However, unlike in the real machine, the rotor is driven by a thin shaft that penetrates through the center of the IGV. Refractive index-matching is used to facilitate optical measurements. Hence, all the blades and housing of the pump are made of transparent acrylic. The working fluid is a mixture of water, sodium iodide and glycerin, which matches the refractive index of acrylic, and the kinematic viscosity of blood. Performance tests have been carried out at speeds ranging from 7000 to 9000 RPM. Trends of the results are consistent with those of the actual machine. While scaled data for the pressure rise across the rotor at different speeds collapse, the total head rise across the entire machine does not. High-resolution 2D PIV measurements with vector spacing of 30μm have been conducted in meridional planes of the rotor passage at several blade orientations. They have been performed at 8000RPM and flowrate of 4.5L/min, consistent with physiological requirements. They show that near the rotor front end, the flow in the tip region is dominated by the tip leakage vortex (TLV), associated blockage effects, and very high turbulence level. The upstream end of this domain remains aligned with the leading edge, implying that blades persistently dissect the remnants of a previous TLVs. Interaction of the hub boundary layer with the blades also generates secondary flows. Downstream of the helical section of the rotor, the axial flow is high near the hub and low in the outer part of the passage due to leakage flow-related blockage. Hub boundary layer separation occurs upstream of the stator, generating a large circumferential vortex that occupies nearly half of the rotor span. The complexity of the flow structure and turbulence in this machine would be a challenge to model.

In the present work, time-dependent responses of Nusselt number, friction coefficient and pressure profiles to the passage of groups of coherent structures along a curved impingement wall, is considered. It is meant to replicate a more realistic picture of the flow. The jet considered belongs to heating applications where the jet flow temperature is higher than that of the impingement wall. The flow was simulated using Large Eddy Simulation with the Dynamic Smagorinsky sub-grid-scale model. The plane jet was forced at frequencies increasing gradually to a maximum of 2200 Hz with an amplitude equal to 30% of the mean jet velocity. The computational domain was divided into 16.5 million hexahedral computational cells whose resolution was assessed based on the turbulence scales. It was found that for low forcing frequencies (e.g., 200Hz), coherent forced primary vortices induced by the pulsations are separated by less organized vortices naturally induced similar to those of the unforced jet. It could be seen that the natural vortices have moderate effects on the boundary layer development on the impingement surface starting at relatively short distances from the stagnation point compared to the forced vortices. Increasing the forcing frequency to 1000Hz reduces the distance separating successive forced vortices causing the pairing phenomenon to occur at a certain distance along the target wall. Increasing the forcing frequency further to 2200Hz makes the pairing phenomenon followed by vortex breakdown to occur at shorter distances along the target wall. The smaller forcing frequencies yield large and strong distant vortices which affect the dynamical field noticeably in conjunction with an important deterioration of heat transfer due to their strong mixing effect and entrainment of cold air from the surroundings. On the other hand, high frequencies generate smaller vortices which are relatively close to each other. Thus, they have a weaker effect allowing the growth of the boundary layer on the target wall up to a distance equal to four times the jet-exit width where the minimum heat transfer is observed. In fact, the small successive vortices form a sort of shield preventing the cold air from the surroundings to reach the target wall until their breakdown.

A Marcellus shale rock fracture was subjected to four shearing steps and at the end of each shearing step CT (computed tomography) scans with resolution of 26.8 μm were obtained. The CT images were used to generate full aperture maps of the fracture configuration at the end of each shearing phase. The pressure drops along the fracture were also measured for different water flow rates through the fracture. The aperture map of the fracture was used to generate the geometry of the fracture for use in numerical simulations. The water flows and pressure drops in the fracture were simulated with different computational methods that included the full Navier-Stokes simulation, Modified Local Cubic Law (MLCL), and Improved Cubic Law (ICL) methods. Full 3-D Navier-Stokes simulation is the most accurate computational approach which was done with use of the ANSYS-Fluent software for each shear step and different flow rates. The MLCL is a 2-D relatively fast method which is commonly used for prediction of transmissivity of fractures. ICL is a 1-D method proposed in this study in which the effects of surface roughness and tortuosity were included in calculation of the effective aperture height of fractures. To provide an understanding of the accuracy of each of these models their predictions were compared with each other and with the experimental data. Also, to examine the effects of resolution of CT scans and the surface roughness on prediction of fractures transmissivity, similar simulations were performed on average aperture maps. Here the fracture of the full resolution data was averaged over 10 × 10 pixels. Comparing the results of the average aperture maps with those of the full maps showed that the lower resolution of CT scans led to underestimation of the fracture pressure drop due to missing the small features of the fracture surfaces and smoothing out their roughness.

Large-scale wind turbine installations are sited using layouts based on site topology, real estate costs and restrictions, and turbine power output. Existing optimization programs have limited capabilities to site multiple turbines and are based on simple geometric turbine wake models, which typically overestimate individual turbine output. Alternatively, complete CFD modeling of entire wind turbine fields requires enormous computational resources, which has led to the development of blade modeling techniques which are combined with CFD field computations. The most promising method, using the actuator line model, typically uses an exponential function to spread blade forces over CFD grid points. In addition, little development work has been performed to determine the optimal grid point density and force spreading radius for these methods. In this paper, we report on our ongoing efforts to develop an advanced actuator line formulation which uses an alternate geometric method for distributing blade forces to the CFD field. Domain and blade force application parameters are currently being developed to determine optimum run time conditions for the new actuator line model. The Actuator line method is implemented using the parallel CFD program, NEK5000. NEK5000 is an advanced Navier Stokes code which uses spectral methods for the spatial discretization, and has been proven to provide high-resolution results with significantly reduced compute resources. A Large Eddy Simulation turbulence model is used. In this paper, we report on our current work using large scale supercomputer resources at the Extreme Science and Engineering Discovery Environment (XSEDE) to perform computational experiments to validate our codes, and perform parametric studies to develop optimum run time parameters. Development and verification work is centered around domain size, grid spacing and clustering, and development of steady state conditions. The parametric studies are underway and are based on investigating various selection volume and force application point settings. Continuing work will compare the new actuator line method with a traditional exponential force distribution model.

Turbulent free jets attracted the focus of many scientists within the past century regarding the understanding of mass- and momentum transport in the turbulent shear field, especially in the near-field and the self-similar region. Recent investigations attempt to understand the intermediate fields, called the mixing transition or ‘the route to self-similarity’. An apparent gap is recognized in light of this mixing transition, with two main conjectures being put forth. Firstly the flow will always asymptotically reach a fully self-similar state if boundary conditions permit. The second proposes partial and local self-similarity within the mixing transition. We address the later with an experimental investigation of the intermediate field turbulence dynamics in a non-confined free jet with a nozzle diameter of 12.7 mm and an outer scale Reynolds number of 15,000. High speed Particle Image Velocimetry (PIV) is used to record the velocity fields with a final spatial resolution of 194 × 194 μm 2 . The analysis focuses on higher order moments and two-point correlations of velocity variances in space and time. We observed local self-similarity in the measured correlation fields. Coherent structures are present within the near-field where the turbulent energy spectrum cascades along a dissipative slope. Towards the transition region, the spectrum smoothly transforms to a viscous cascade, as it is commonly observed in the self-similar region.

A combined vortex identification and Proper Orthogonal Decomposition (POD) analysis is applied to high-resolution Particle Image Velocimetry (PIV) measurements of a turbulent flow past an open shallow cavity. The PIV measurements, at a cavity depth based Reynolds number of 42,000, capture the flow structure and turbulence, upstream, over, and downstream an open cavity having a length-to-depth ratio of four. Vorticity and second invariant Q of the velocity gradient tensor analysis are used to identify the vortical structures and the overall flow field features. POD analysis is applied to the vorticity and Q fields to identify the most energetic vortical structures and flow features. The results demonstrate the superiority of the combined Q-criterion and POD analysis in identifying distinct vortical structures and their evolution.

Liquid metal infiltration, or liquid method infusion, consists of impregnating porous media composed of woven, ceramic particles, or fibers with a molten metal matrix, which fills the pores and occupies the void space within. Understanding the infiltration process is crucial to optimize the properties of the recently formed material and avoid or minimize the formation of fabrication defects. Given the fact that the flow of molten metal differs from organic flows, since molten metal possess a higher interface energy than organic flows, and modifies the wetting dynamics of the molten metal over surfaces, creating a flow driven by capillary and viscous forces. In addition, flow through porous media presents an extraordinary challenge to simulate efficiently, due to the presence of multiple scales far apart participating in the governing dynamics. For this reason, an in-house pore network simulator (EXPNS) was used. EXPNS was designed on a next generation computing framework using Sandia National Lab’s Trilinos and Kokkos library to perform high-resolution computing to generate data for the infiltration model and improve the general understanding of this process.

The development and validation of mechanistic boiling heat transfer models has been a focal point in the efforts to improve the efficiency and profitability of power generation systems, e.g. nuclear reactors. The primary goal of these models is improving the accuracy of boiling heat transfer simulations and reducing the uncertainty margins that affect both the design and the safety of a system. However, the emergence of these models has also stimulated the need for high-fidelity experiments and experimental data for validation and verification. In this work we present first-of-a-kind data of heat flux partitioning in boiling heat transfer, obtained using cutting-edge diagnostics and post-processing techniques. A HSV camera was used to visualize the boiling surface at 10,000 frames per second with simultaneous front and side views of the two-phase flow. A high-speed IR camera was used to capture the 2-D radiative signal from the boiling surface to visualize bubble nucleation, growth and detachment at a 115 μm/pixel resolution at 2,500 frames per second. A coupled radiation-conduction calibration model was used to calibrate the IR data and extract the full local temperature and heat flux distributions on the boiling surface, which enable a direct measurement of the partitioned heat fluxes. Here we report the results of investigations performed in flow boiling conditions with a mass flux of 500 kg/m 2 /s, at atmospheric pressure and 10 K of subcooling. These data will be leveraged to inform the development and validation of the next generation of mechanistic boiling heat transfer models, to be used in Computational Fluid Dynamics (CFD) codes for the design and the safety analysis of nuclear reactors.

DNS coupled with a Point-Particle based model (PP) is used to study and predict particle-turbulence interactions in an open channel flow at Reynolds number of 811 (based on the friction velocity) corresponding to the experimental observations of [Righetti & Romano, JFM 2004]. Large particles of diameter 200 microns (8.1 in wall units) with average volume loading on the order of 0.001 are simulated using four-way coupling with closure models for drag, added mass, lift, pressure, and inter-particle/particle-wall collision forces. The point-particle model is able to accurately capture the effect of particles on the fluid flow in the outer layer where particles are under resolved. However, the dynamical interaction of particle-turbulence is under predicted in the near wall region where particles size are much larger than Kolmogorov scale and grid resolution in wall-normal direction, but smaller in both stream and span wise directions. It is conjectured that due to the large size particles compared to the Kolmogorov length scale near the bed, the effect of disturbances and deflections in the flow due to presence of such large particles is not captured using Lagrangian Point-Particle approach. For this configuration, the point-particle model is not appropriate in the near wall region and a hybrid resolved particle approach may be necessary.

A new high-order sharp treatment of mixed cell solutions is introduced. The method is based on the reconstructed discontinuous Galerkin (rDG) scheme, in which we embed the known interfacial jump conditions as additional (physics-based) constraints in the least-squares reconstruction of the high-order degrees of freedom for multiple solution fields in the interfacial (mixed) cells. The approach allows us to avoid direct differentiation across multi-material interfaces, thus providing a robust and high-order accurate solution procedure. To track the interface dynamics, the method is combined with a hybrid of the level set (LS) and the front tracking (FT) methods — the so-called “Marker-Re-Distancing” (MRD) approach. As a fluid solver, we utilize the recently developed fully-implicit reconstructed discontinuous Galerkin method, developed for efficient and high-order resolution of all-speed compressible fluid dynamics. The new sharp physics-based reconstruction is incorporated into the overall Newton-Krylov-based solution procedure, with the residual of the mixed cell representing conservation of the mass, momentum and energy for the mixture. Thus, upon a convergence of the non-linear iterations, our treatment of interfacial cells is conservative.

We focus on the simulation of shock-driven material mixing powered by flow instabilities dependent on initial conditions (IC) at the material interfaces. Beyond complex multi-scale resolution issues of shocks and variable density turbulence, we must address the equally difficult problem of predicting flow transition promoted by energy deposited at the interfacial layers during the shock-interface interactions. Transition involves IC-dependent, large-scale coherent-structure dynamics capturable by a large eddy simulation (LES) strategy, but not by unsteady Reynolds-Averaged Navier-Stokes (URANS) approaches based on equilibrium developed turbulence assumptions and single-point-closure modeling. On the engineering end of computations, reduced-dimensionality (1D/2D) versions of such URANS tend to be preferred for faster turnaround in full-scale configurations. With suitable initialization around each transition, URANS can be used to simulate the subsequent near-equilibrium weakly turbulent flow. We demonstrate 3D state-of-the-art URANS performance around one such (reshock) transition — in the context of a sequential LES/URANS strategy.

A research wind turbine of one meter diameter was designed for the UNH Flow Physics Facility (FPF), a very large flow physics quality turbulent boundary layer wind tunnel (W 6m, H 2.7m, L 72m), which provides excellent spatial and temporal resolution, low flow blockage and allows measurements of turbine wakes far downstream due its long fetch. The initial turbine design was carried out as an aero-servo model of the NREL 5MW reference turbine, with subsequent modifications to both the hub to accommodate blade mounting and pitch-adjustment, and increases in model blade chord to achieve sufficiently high Reynolds numbers. A trade-off study of turbine design parameters in scale space was conducted. Several candidate airfoil profiles were evaluated numerically with the goal to reach Reynolds-number independence in turbine performance in the target operating range. The model turbine will achieve Reynolds numbers based on blade chord, an important consideration for airfoil performance and near-wake evolution, greater than 100,000, and Reynolds numbers based on turbine diameter, important for far-wake transport, on the order of 1,000,000. The blockage ratio is less than 5% based on swept area. A motor and controller combination was implemented that allows to precisely prescribe the turbine tip-speed ratio (at maximum power coefficient for optimum blade chord), which can remain stable and absorb the generated electric power for long periods of time. The turbine nacelle was designed with a blade mounting mechanism which allows for precise manual adjustment of blade pitch angle, while allowing for future implementation of actuated pitch control. The O (1 m ) turbine scale is viewed as a cost-effective compromise between size, driven by the need for sufficiently high Reynolds number, and the need for detailed measurements for significant distances downstream of the turbine under controlled conditions.

The helical coil steam generator (HCSG) is a compact heat exchanger that can have high heat transfer even when the pressure drop is low. This makes it advantageous in small modular reactors and high-temperature reactor designs. In order to investigate the fluid phenomena around these helical banked tubes, a test section was built at Texas A&M University to represent flow across two half-rods within HCSG. This study focuses on the validation of large eddy simulation (LES) for this particular geometry. Pressure tap and particle image velocimetry (PIV) measurements have been recorded at an inlet Reynolds number of 8643, and both mean and fluctuating data is compared with the numerical results. The highly scalable spectral-element code Nek5000 has been used to produce the LES calculations. First, simulations of varying polynomial order expansions are made to determine the spatial resolution required to capture the turbulent scales. Then, simulations with different inlet conditions are compared with experimental data. The pressure drop shows good agreement with pressure tap measurements while velocity shows similar characteristics with PIV. Furthermore, the components of the Reynolds stresses and modes from proper orthogonal decomposition have been developed to validate the physics captured.