The importance of the turbulence closure to the modeling accuracy of the partially-averaged Navier–Stokes equations (PANS) is investigated in prediction of the flow around a circular cylinder at Reynolds number of 3900. A series of PANS calculations at various degrees of physical resolution is conducted using three Reynolds-averaged Navier–Stokes equations (RANS)-based closures: the standard, shear-stress transport (SST), and turbulent/nonturbulent (TNT) k–ω models. The latter is proposed in this work. The results illustrate the dependence of PANS on the closure. At coarse physical resolutions, a narrower range of scales is resolved so that the influence of the closure on the simulations accuracy increases significantly. Among all closures, PANS–TNT achieves the lowest comparison errors. The reduced sensitivity of this closure to freestream turbulence quantities and the absence of auxiliary functions from its governing equations are certainly contributing to this result. It is demonstrated that the use of partial turbulence quantities in such auxiliary functions calibrated for total turbulent (RANS) quantities affects their behavior. On the other hand, the successive increase of physical resolution reduces the relevance of the closure, causing the convergence of the three models toward the same solution. This outcome is achieved once the physical resolution and closure guarantee the precise replication of the spatial development of the key coherent structures of the flow.

In this paper, capabilities of state-of-the-art computational fluid dynamics (CFD) tools in the prediction of the flow-field around a multihull catamaran advancing in straight ahead motion at nonzero drift angles are investigated. CFD estimations have been provided by three research institutes by using their in-house codes: CNR-INM using Xnavis, IIHR using CFDShip-Iowa, and CNRS/ECN using ISIS. These allowed an in-depth comparison between different methodologies, such as structured overlapping grids versus unstructured grid, different turbulence models and detached eddy simulations (DES) approaches, and level-set (LS) versus volume of fluid (VoF). The activities were pursued within the NATO AVT-183 group “reliable prediction of separated flow onset and progression for air and sea vehicles,” aimed at the assessment of CFD predictions of large three-dimensional separated flows. Comparison between estimations is provided for both integral and local quantities, and for wave-induced vortices. Validation is reported by comparison against the available experimental fluid dynamics (EFD) data. Generally, all the simulations are able to capture the main features of the flow field; grid resolution effects are dominant in the onset phase of coherent structures and turbulence model affects the dynamic of the vortices. Hydrodynamic loads are in agreement between the submissions with standard deviation of about 3.5% for the resistance prediction and about 7% for lateral force and yaw moment estimation. Wave-induced vortices are correctly captured by both LS and VoF approaches, even if some differences have been highlighted, LS showing well-defined and long life vortices.

A contact angle control algorithm is developed and implemented in the multiphase interface tracking flow solver— phasta . The subgrid force model is introduced to control the evolving contact angle. The contact angle force is applied when the current contact angle deviates from the desired value (or range of values) and decreases to zero when it reaches the desired value. The single bubble departure simulation and the capillary flat plates simulation are performed for verification purpose. The numerical results are compared with the analytical solution with good agreement. The mesh resolution sensitivity analysis and parametric study are conducted for both simulations. Coupled with the other existing capabilities in phasta like evaporation and condensation algorithm, the contact angle control algorithm will allow us to investigate the boiling phenomenon in various conditions with lower cost (by utilizing localized mesh refinement for bubble growth region) compared to uniformly refined structured meshes and in engineering geometries.

Large eddy simulation (LES) is conducted for the flow over the shell side of a helical coil steam generator (HCSG) heat exchanger. Simulations are conducted on a simplified experimental test section that represents a one-column region of the helical coils using half-rods. Although the rods are wall-bounded, the flow still exhibits the turbulent characteristics and fluctuations from vortex shedding that one would expect from crossflow around a cylinder. The spectral element, computational fluid dynamics (CFD) code Nek5000, is used to capture the physics, and the results are compared with particle image velocimetry (PIV) measurements. In order to ensure that the turbulence is resolved, analysis is conducted by using the Taylor length scales and normalized wall distance. Sensitivity to the inlet boundary conditions (BCs) and the spatial discretization for different polynomial order solutions are also studied, finding only minor differences between each case. Pressure drop and velocity statistics show reasonable agreement with PIV. Proper orthogonal decomposition (POD) analysis reveals that the primary modes are similar between experiment and simulation, although the LES predicts higher turbulent kinetic energy than does PIV. Overall, the study establishes the resolution and resources required in order to conduct a high-fidelity simulation over 12 helical rods.

An efficient large-eddy simulation (LES) approach is investigated for laminar-to-turbulent transition in boundary layers. This approach incorporates the boundary-layer stability theory. Primary instability and subharmonic perturbations determined by the boundary-layer stability theory are assigned as forcing at the inlet of the LES computational domain. This LES approach reproduces the spatial development of instabilities in the boundary layer, as observed in wind tunnel experiments. Detailed linear growth and nonlinear interactions that lead to the H-type breakdown are well captured and compared well to previous direct numerical simulation (DNS). Requirements in the spatial resolution in the transition region are investigated with connections to the resolution in turbulent boundary layers. It is shown that the subgrid model used in this study is apparently dormant in the overall transitional region, allowing the right level of the growth of small-amplitude instabilities and their nonlinear interactions. The subgrid model becomes active near the end of the transition where the length scales of high-order instabilities become smaller in size compared to the given grid resolution. Current results demonstrate the benefit of the boundary-layer forcing method for the computational cost reduction.

We develop a new high-order numerical method for continuum simulations of multimaterial phenomena in solids exhibiting elastic–plastic behavior using the diffuse interface numerical approximation. This numerical method extends an earlier single material high-order formulation that uses a tenth-order high-resolution compact finite difference scheme in conjunction with a localized artificial diffusivity (LAD) method for shock and contact discontinuity capturing. The LAD method is extended here to the multimaterial formulation and is shown to perform well for problems involving shock waves, material interfaces and interactions between the two. Accuracy of the proposed approach in terms of formal order (eighth-order) and numerical resolution is demonstrated using a suite of test problems containing smooth solutions. Finally, the Richtmyer–Meshkov (RM) instability between copper and aluminum is simulated in two-dimensional (2D) and a parametric study is performed to assess the effect of initial perturbation amplitude and yield stress.

We have studied the influence of a tangential blowing jet in dynamic stall of a NACA0012 airfoil at Reynolds number of 1 × 10 6 , for active flow control (AFC) purposes. The airfoil was oscillating between angles of attack (AOA) of 5 and 25 deg about its quarter-chord with a sinusoidal motion. We have utilized computational fluid dynamics to investigate the impact of jet location and jet velocity ratio on the aerodynamic coefficients. We have placed the jet location upstream of the counter-clockwise (CCW) vortex which was formed during the upstroke motion near the leading-edge; we have also considered several other locations nearby to perform sensitivity analysis. Our results showed that placing the jet slot within a very small range upstream of the CCW vortex had tremendous effects on both lift and drag, such that maximum drag was reduced by 80%. There was another unique observation: placing the jet at separation point led to an inverse behavior of drag hysteresis curve in upstroke and downstroke motions. Drag in downstroke motion was significantly lower than upstroke motion, whereas in uncontrolled case the converse was true. Lift was significantly enhanced during both upstroke and downstroke motions. By investigating the pressure coefficients, it was found that flow control had altered the distribution of pressure over the airfoil upper surface. It caused a reduction in pressure difference between upper and lower surfaces in the rear part, while substantially increased pressure difference in the front part of the airfoil.

Despite recent interests in complex fluid–structure interaction (FSI) problems, little work has been conducted to establish baseline multidisciplinary FSI modeling capabilities for research and commercial activities across computational platforms. The current work investigates the fluid modules of contemporary FSI methodologies by solving a purely fluid problem at low Reynolds numbers to improve understanding of the fluid dynamic capabilities of each solver. By incorporating both monolithic and partitioned solvers, a holistic comparison of computational accuracy and time-expense is presented between lattice-Boltzmann methods (LBM), coupled Lagrangian–Eulerian (CLE), and smoothed particle hydrodynamics (SPH). These explicit methodologies are assessed using the classical square lid-driven cavity for low Reynolds numbers (100–3200) and are validated against an implicit Navier–Stokes solution in addition to established literature. From an investigation of numerical error associated with grid resolution, the Navier–Stokes solution, LBM, and CLE were all relatively mesh independent. However, SPH displayed a significant dependence on grid resolution and required the greatest computational expense. Throughout the range of Reynolds numbers investigated, both LBM and CLE closely matched the Navier–Stokes solution and literature, with the average velocity profile error along the generated cavity centerlines at 1% and 4%, respectively, at Re = 3200. SPH did not provide accurate results whereby the average error for the centerline velocity profiles was 31% for Re = 3200, and the methodology was unable to represent vorticity in the cavity corners. Results indicate that while both LBM and CLE show promise for modeling complex fluid flows, commercial implementations of SPH demand further development.

An algorithm to prevent or delay bubble coalescence for the level set (LS) method is presented. This novel algorithm uses the LS method field to detect when bubbles are in close proximity, indicating a potential coalescence event, and applies a repellent force to simulate the unresolved liquid drainage force. The model is introduced by locally modifying the surface tension force near the liquid film drainage area. The algorithm can also simulate the liquid drainage time of the thin film by controlling the length of time the increased surface tension has been applied. Thus, a new method of modeling bubble coalescence has been developed. Several test cases were designed to demonstrate the capabilities of the algorithm. The simulations, including a mesh study, confirmed the abilities to identify and prevent coalescence as well as implement the time tracking portion, with an additional 10–25% computational cost. Ongoing tests aim to verify the algorithm's functionality for simulations with different flow conditions, a ranging number of bubbles, and both structured and unstructured computational mesh types. Specifically, a bubble rising toward a free surface provides a test of performance and demonstrates the ability to consistently prevent coalescence. In addition, a two bubble case and a seven bubble case provide a more complex demonstration of how the algorithm performs for larger simulations. These cases are compared to much more expensive simulations capable of resolving the liquid film drainage (through very high local mesh resolution) to investigate how the algorithm replicates the liquid film drainage process.

The compressible Rayleigh–Taylor (RT) instability is studied by performing a suite of large eddy simulations (LES) using the Miranda and Ares codes. A grid convergence study is carried out for each of these computational methods, and the convergence properties of integral mixing diagnostics and late-time spectra are established. A comparison between the methods is made using the data from the highest resolution simulations in order to validate the Ares hydro scheme. We find that the integral mixing measures, which capture the global properties of the RT instability, show good agreement between the two codes at this resolution. The late-time turbulent kinetic energy and mass fraction spectra roughly follow a Kolmogorov spectrum, and drop off as k approaches the Nyquist wave number of each simulation. The spectra from the highest resolution Miranda simulation follow a Kolmogorov spectrum for longer than the corresponding spectra from the Ares simulation, and have a more abrupt drop off at high wave numbers. The growth rate is determined to be between around 0.03 and 0.05 at late times; however, it has not fully converged by the end of the simulation. Finally, we study the transition from direct numerical simulation (DNS) to LES. The highest resolution simulations become LES at around t/τ ≃ 1.5. To have a fully resolved DNS through the end of our simulations, the grid spacing must be 3.6 (3.1) times finer than our highest resolution mesh when using Miranda (Ares).

Although coaxial airblast primary atomization has been studied for decades, relatively little attention has been given to three-stream designs; this is especially true for transonic self-pulsating injectors. Herein, the effects of nozzle geometry, grid resolution, modulation, and gas flow rate on the acoustics and spray character within an industrial scale system were investigated computationally using axisymmetric (AS) and three-dimensional (3D) models. Metrics included stream pressure pulsations, spray lift-off, spray angle, and primary droplet length scale, along with the spectral alignment among these parameters. Strong interactions existed between geometry and inner gas (IG) feed rate. Additionally, inner nozzle retraction and outer stream meeting angle were intimately coupled. Particular attention was given to develop correlations for various metrics versus retraction; one such example is that injector flow capacity was found to be linearly proportional to retraction. Higher IG flows were found to widen sprays, bringing the spray in closer to the nozzle face, and reducing droplet length scales. Substantial forced modulation of the IG at its dominant tone did not strongly affect many metrics. Incompressible 3D results were similar to some of the AS results, which affirmed the predictive power by running AS simulations as surrogates. Lastly, normalized droplet size versus normalized distance from the injector followed a strikingly similar trend as that found from prior two-fluid air-slurry calibration work.

The investigation of the rotating stall phenomenon appearing in the HYDRODYNA pump-turbine reduced scale model is carried out by performing a large-scale large eddy simulation (LES) computation using a mesh featuring approximately 85 × 10 6 elements. The internal flow is computed for the pump-turbine operated at 76% of the best efficiency point (BEP) in pumping mode, for which previous experimental research evidenced four rotating stall cells. To achieve an adequate resolution near the wall, the Reynolds number is decreased by a factor of 25 than that of the experiment, by assuming that the flow of our interest is not strongly affected by the Reynolds number. The computations are performed on the supercomputer PRIMEHPC FX10 of the University of Tokyo using the overset finite-element open source code FrontFlow/blue with the dynamic Smagorinsky turbulence model. It is shown that the rotating stall phenomenon is accurately simulated using the LES approach. The results show an excellent agreement with available experimental data from the reduced scale model tested at the EPFL Laboratory for hydraulic machines. The number of stall cells as well as the propagation speed agree well with the experiment. Detailed investigations on the computed flow fields have clarified the propagation mechanism of the stall cells.

A study was performed to investigate the interesting observation that when using an Eulerian–Eulerian model to simulate a bubble column flow in two dimensions, accuracy did not always increase with increasing grid resolution. A correlation was found between the characteristic bubble diameter numerically specified and grid size, which identified a threshold where results lose physical meaning. An ideal relationship between grid size and bubble diameter was determined to optimize grid resolution and retain accuracy. The two-dimensional (2D) Eulerian model was validated using the experimental data of Rampure et al. (2003, “Modelling of Gas-Liquid/Gas-Liquid-Solid Flows in Bubble Columns: Experiments and CFD Simulations,” Can. J. Chem. Eng., 81 (3–4), pp. 692–706). Further studies demonstrated that grid resolution could be increased to improve the numerical accuracy for three-dimensional (3D) simulations of the bubble column. The novel contributions of this study will show that the ratio of bubble diameter-to-grid cell size should equal 1 / 2 for the 2D simulations.

A fully developed turbulent particle-gas flow in a rectangular horizontal channel 100 × 10 × 4000 mm 3 is disclosed with high spatial resolution two-dimensional (2D) particle image velocimetry (PIV). The objective is to increase the knowledge of the mechanisms behind alterations in turbulent characteristics when adding two sets of relatively large solid spherical particles with mean diameters of 525 and 755 μ m and particle size distributions of 450–600 and 710–800 μ m, respectively. Reynolds numbers are 4000 and 5600 and relatively high volume fraction of 5.4 × 10 −4 and 8.0 × 10 −4 are tested. Both the near wall turbulent boundary layer flow and outer core flow are considered. Results show that the carrier phase turbulent intensities increase with the volume fraction of the inertial particles. The overall mean flow velocity is affected when adding the particles but only to a minor extent. Near the wall, averaged velocity decreases while fluctuating velocity components increase when particles are added to the flow. Quadrant analysis shows the importance of sweep near the wall and ejection events in the region defined by y + > 20. In conclusion, high inertia particles can enhance turbulence even at relatively low particle Reynolds number <90. In the near bottom wall region, particles tend to be a source of instability reflected as enhancement in rms values of the normal velocity component.

Methods to determine the pressure field of vortical flow from three-dimensional (3D) volumetric velocity measurements (e.g., from a TSI V3V TM system) are discussed. The boundary pressure was determined where necessary using the unsteady Bernoulli equation for both line integration and pressure Poisson equation methods. Error analysis using computational fluid dynamics (CFD) data was conducted to investigate the effects of spatial resolution, temporal resolution, and velocity error levels. The line integration method was more sensitive to temporal resolution, while the pressure Poisson equation method was more sensitive to boundary flow conditions. The latter was generally more suitable for V3V TM velocity measurements.

This paper deals with the computational predictability of aerodynamic losses in a turbine nozzle guide vane (NGV) flow. The paper shows that three-dimensional (3D) computations of Reynolds-Averaged Navier Stokes (RANS) equations have the ability to adequately represent viscous losses in the presence of laminar flows, transitional regions, and fully turbulent flow areas in the NGV of an high pressure (HP) turbine stage. The Axial Flow Turbine Research Facility (AFTRF) used for the present experimental results has an annular NGV assembly and a 29-bladed HP turbine rotor spinning at 1330 rpm. The NGV inlet and exit Reynolds numbers based on midspan axial chord are around 300,000 and 900,000, respectively. A general purpose finite-volume 3D flow solver with a shear stress transport (SST) k–ω turbulence model is employed. The current computational study benefits from these carefully executed aerodynamic experiments in the NGV of the AFTRF. The grid independence study is performed with static pressure coefficient distribution at the midspan of the vane and the total pressure coefficient at the NGV exit. The effect of grid structure on aerodynamic loss generation is emphasized. The flow transition effect and the influence of corner fillets at the vane–endwall junction are also studied. The velocity distributions and the total pressure coefficient at the NGV exit plane are in very good agreement with the experimental data. This validation study shows that the effect of future geometrical modifications on the turbine endwall surfaces will be predicted reasonably accurately. The current study also indicates that an accurately defined turbine stage geometry, a properly prepared block-structured/body-fitted grid, a state-of-the-art transitional flow implementation, inclusion of fillets, and realistic boundary conditions coming from high-resolution turbine experiments are all essential ingredients of a successful turbine NGV aerodynamic loss quantification via computations. This validation study forms the basis for the successful future generation of nonaxisymmetric endwall surface modifications in AFTRF research efforts.

Flow behavior in complex three-dimensional (3D) microscale domains is the key in the development of microcirculatory pathologies and the design of 3D microfluidics. While numerical simulations are common practice for the derivation of velocity fields in such domains, they are limited to known geometries. Current experimental methods such as micron-scale particle tracing comprise of intricate algorithmic approaches for the accurate tracing of numerous particles in a dense moving liquid suspension and are fundamentally limited in resolution to the finite size of the interrogated steps. Here, we introduce 3D streamlines image velocimetry (3D-SIV), a method to derive fluid velocity fields in arbitrary resolution for fully developed laminar flow in 3D geometries. Our approach utilizes 3D geometrical fitting and superimposed Delaunay triangulation to reconstruct streamtubes and to trace their volumetric changes. Our algorithm has applications in out-of-plane velocimetries, which we demonstrate in a 3D dilated curved geometry and in an ascending aorta. The 3D-SIV can be applied for high-resolution derivation of velocity fields in microcirculatory pathologies and to 3D microfluidic circuits, extending the potential of out-of-plane velocimetries to complex geometries and arbitrary resolution.

This paper presents an investigation into the use of a moving mesh algorithm for solving unsteady turbulent mixing problems. The growth of a shock induced mixing zone following reshock, using an initial setup comparable to that of existing experimental work, is used to evaluate the behavior of the numerical scheme for single-mode Richtmyer–Meshkov instability (SM-RMI). Subsequently the code is used to evaluate the growth rate for a range of different initial conditions. The initial growth rate for three-dimensional (3D) SM Richtmyer–Meshkov is also presented for a number of different initial conditions. This numerical study details the development of the mixing layer width both prior to and after reshock. The numerical scheme used includes an arbitrary Lagrangian–Eulerian grid motion which is successfully used to reduce the mesh size and computational time while retaining the accuracy of the simulation results. Varying initial conditions shows that the growth rate after reshock is independent of the initial conditions for a SM provided that the initial growth remains in the linear regime.

This paper presents a hybrid compressible–incompressible approach for simulating the Richtmyer–Meshkov instability (RMI) and associated mixing. The proposed numerical approach aims to circumvent the numerical deficiencies of compressible methods at low Mach (LM) numbers, when the flow has become essentially incompressible. A compressible flow solver is used at the initial stage of the interaction of the shock wave with the fluids interface and the development of the RMI. When the flow becomes sufficiently incompressible, based on a Mach number prescribed threshold, the simulation is carried out using an incompressible flow solver. Both the compressible and incompressible solvers use Godunov-type methods and high-resolution numerical reconstruction schemes for computing the fluxes at the cell interfaces. The accuracy of the model is assessed by using results for a two-dimensional (2D) single-mode RMI.

Nested numerical simulations of ionospheric plasma density structures associated with nonlinear evolution of the Rayleigh–Taylor (RT) instability in equatorial spread F (ESF) are presented. The numerical implementation of the nested model uses a spatial discretization with a C grid staggering configuration where normal velocities of ions and electrons are staggered one-half grid length from the density of charged particles. The advection of charged particles is computed with a fifth order accurate in space weighted essentially nonoscillatory (WENO) scheme. The continuity equation is integrated using a third-order Runge–Kutta (RK) time integration scheme. The equation for the electric potential is solved at each time step with a multigrid method. For the limited area and nested simulations, the lateral boundary conditions are treated via implicit relaxation applied in buffer zones where the density of charged particles for each nest is relaxed to that obtained from the parent domain. The high resolution in targeted regions offered by the nested model was able to resolve secondary RT instabilities, and to improve the resolution of the primary RT bubble compared to the coarser large domain model. The computational results are validated by conducting a large domain simulation where the resolution is increased everywhere.