Search

Results of a coordinated research effort on unidirectional turbulent flows over canonical barchan dunes at high Reynolds number are presented. Large-eddy simulations (LES) and experiments are conducted under inertial-dominated flow conditions, thus capturing dynamics essential to fully understanding processes responsible for observed patterns of asymmetric erosion and migration. A series of dune field topographies have been considered wherein a small “impactor” dune is positioned at a series of positions upflow of a large “parent” dune, from a spanwise-offset position; this interaction constitutes a so-called “offset interaction” Kocurek and Ewing [14]. The small dune is geometrically similar, but one-eighth the volume of the large dune. Since migration rate is inversely proportional to volume, the prescribed volumetric ratio ensures that the synthetic topographies replicate instantaneous configurations exhibited during actual dune interactions in a laboratory or natural setting. In this sense, the static dune configurations provide a means to understand flow processes responsible for patterns of erosion and deposition that induce interactions. Experimental measurement and LES are both used to study these configurations, with strong agreement reported between resultant datasets. We report that flow channeling in the interdune space between the upflow and downflow dune induces a mean flow heterogeneity – termed “wake veering” – in which the location of maximum momentum deficit in the dune wake is spanwise displaced. We also report elevated turbulent stresses in the channeling region. These results are used to explain the mechanisms responsible for simultaneous merging and ejection between the small and large dunes. Finally, spatial distributions of surface stress from LES have been used to identify locations of elevated erosion and, therefore, to predict bedform migration patterns. Results show that locations of minimal erosion – whether associated with upflow sheltering or with vanishing spatial gradients of dune height – constitute spatial “junctions” of coalescing, proximal dunes. In the absence of upwind sheltering, or in locations with the least upwind sheltering, turbulent mixing facilitates downward fluxes of high momentum fluid. This augments downwind dune erosion, cumulatively accelerating the migration of downwind dunes. [ABSTRACT FROM AUTHOR]

The paper presents aerodynamic and fluid–structure interaction (FSI) simulations of two back-to-back 5 MW horizontal-axis wind turbines (HAWTs) at full scale and with full geometrical complexity operating in a stably-stratified atmospheric boundary layer (ABL) flow. The numerical formulation for stratified incompressible flows is based on the ALE-VMS methodology, and is coupled to a Kirchhoff–Love thin-shell formulation employed to model the wind-turbine structure. A multi-domain method (MDM) is adopted for computational efficiency. In the simulations presented the wind turbines are positioned one behind the other at a distance of four rotor diameters, which results in a noticeable power production loss for the downstream turbine due to the wake velocity deficit of the upstream turbine. The importance of including FSI coupling in the modeling to better predict the unsteady aerodynamic loads acting on the wind-turbine blades is also highlighted. [ABSTRACT FROM AUTHOR]

This paper is dealing with gradient-based optimization in fluid mechanics, in the presence of uncertainties, by focusing on turbulent flows governed by the RANS equations. Uncertainty Quantification (UQ) is performed using the First-Order Second-Moment (FOSM) method and the more accurate Adjoint-assisted (non-intrusive) Polynomial Chaos Expansion (APCE); in the latter, first-order derivatives of the Quantity of Interest (QoI) with respect to the uncertain variables, provided by the adjoint method, are used to reduce the cost for computing the polynomial coefficients. Both methods compute the first two statistical moments (mean value and standard deviation of the QoI) and the objective function to be optimized is their weighted sum. Gradient-based optimization with such an objective function requires mixed derivatives of the QoI with respect to the design and uncertain variables. In adjoint-based optimization, one way to compute these derivatives is by solving the systems of PDEs resulting from the differentiation of the flow and adjoint equations with respect to the uncertain variables, at a cost that scales with their number. To reduce the CPU cost, the objective function gradient can be expressed in terms of the projection of the mixed derivatives' matrix onto vectors. Such projections are herein utilized with both the aforementioned UQ methods, in the context of robust design optimization. Firstly, the projected FOSM method, presented for laminar flows in a previous article by the same group, is extended to turbulent flows solved using the Spalart–Allmaras model by including its adjoint. Then, the projected APCE method is developed, programmed and tested for the first time in the literature. In both pFOSM and pAPCE, the cost for computing the projected matrix of mixed derivatives does not scale with the number of either the design or the uncertain variables. Both methods are implemented within the adjointOptimisation library of OpenFOAM, which makes use of the continuous adjoint method, and are demonstrated in the aerodynamic shape optimization of airfoils, in laminar and turbulent flows. • Novel projected Adjoint-based PCE for gradient-based optimization with uncertainties. • Extension of pFOSM and pAPCE to turbulent flows by using adjoint to turbulence model. • Problem reformulation to make matrix–vector products appear. • Less collocation points for regression-based PCE due to adjoint. • Assessment of cost reduction of projected (compared to standard) FOSM, APCE. [ABSTRACT FROM AUTHOR]

Abstract: In most fluid mechanic applications, the dominant high energy content POD modes are generally well converged justifying the great use of POD to extract the associated large scale flow structures of a turbulent flow. However for particular studies related to the characterization of smaller flow structures, it is necessary to take into account high order POD modes. In this paper based on a particular numerical database available on a finite-discrete spatio-temporal domain, the sensibility of POD modes to both spatial discretization and independent temporal samples is highlighted as a function of integral time and space scale content. A systematic procedure is then proposed in order to verify the spatial convergence of POD modes for a given number of points per integral space scale and per integral time scale and also for a given number of uncorrelated samples that is the number of integral time scale contained in the original signal. The convergence of Extended POD modes is also verified by an error threshold definition that permits to avoid any misinterpretation of the results. A new extension of the EPOD procedure based on the snapshot POD technique is then proposed aiming at using only the converged EPOD modes to characterize the correlated part of the signal. [Copyright &y& Elsevier]

Abstract: Numerous direct numerical simulations of soap film flows yield a large amount of data on two-dimensional turbulence. The analysis of such data is very sensitive to the analysis tools and the way they are used. On the one hand, some of the possible errors obtained by a misuse of the analysis tools are reported. On the other hand, a rigorous use of wavelet packets analysis reveals surprising results that slightly differ from the KLB theory for the flow considered. A careful wavelet packets filtering and a good computation of the energy and enstrophy fluxes show the role of the solid rotation vortices and the vorticity filaments to both the inverse energy cascade and the direct enstrophy cascade. [Copyright &y& Elsevier]

Abstract: A computational fluid dynamics (CFD) model is developed to better understand the complex flow field inside a free surface fish bypass constructed at Rocky Reach Dam. This facility consists of two identical parallel channels, with fish screens on the side walls of each channel, and a pump station recirculating 96% of incoming flows into the forebay. The model is based on the Reynolds-averaged Navier–Stokes (RANS) equations, with a standard κ–ε turbulence model. The volume of fluid (VOF) method is used to predict free surface elevations. A proportional controller is implemented in the model to achieve a target flow rate at the pump exits. The pressure drop in the fish screens is modeled using porous media. Quantitative validation and visualization of the flow field characteristics indicate that CFD modeling may be a useful tool for fish passage design. [Copyright &y& Elsevier]

Abstract: The present study will give rise to better understanding of the airflow behaviour in a ceiling-slot ventilated enclosure partially filled with slotted porous boxes under isothermal conditions. Boxes were filled randomly with spheres of the same diameter. Laser Doppler velocimetry (LDV) measurements and computational fluid dynamics (CFD) predictions were carried out to characterize air velocity inside packed vented boxes. An original approach was also developed to evaluate the internal macroscopic velocities within the boxes by means of thermal sphere-shaped probes. LDV measurements and CFD predictions related to the main turbulent flow developed by the jet above the boxes were presented in the part I. Special attention was paid to the strong aerodynamic interactions observed on the top faces of the slotted boxes, between the external main flow and the return flow which predominate inside the boxes. Numerical and experimental data make it possible to evaluate the air ventilation levels and their heterogeneity for the different boxes. The RSM turbulence model gives reasonable agreement with experiments. [Copyright &y& Elsevier]

Abstract: In this paper, a simplified 3-D algebraic slip mixture (ASM) model was introduced to obtain the numerical solution in sand–water slurry flow. In order for the study to obtain the precise numerical solution in fully developed turbulent flow, the RNG k–ε turbulent model was used with the algebraic slip mixture model. An unstructured (block-structured), non-uniform grid was chosen to discretize the entire computation domain, and a control volume finite difference method (CVFDM) was applied to solve the governing equations. The mean pressure gradients from the numerical solution in the fully developed turbulent flow were compared with the authors’ experimental data and that in the open literature to validate the simulation results. The solutions were found to be in good agreement with the experimental data. The numerical investigations have focused mainly on illustrating and comparing the developing processes of volume fraction and density distributions, mean velocity profiles, and mean skin friction coefficient distributions in the entrance region. It was found that the difference of friction factors between single- and double-species slurries increases along flow direction in most part of the entrance region. [Copyright &y& Elsevier]

Abstract: This paper describes an automatic local grid adaptation procedure driven by an evaluation of the differential residuals of the RANS equations computed using a higher-order reconstruction operator. A suitable data structure is developed for the local mesh adaptation process to be flexible and low CPU time consuming. The whole procedure is designed in the framework of finite-volume methods on unstructured grids. To avoid the appearance of ill-conditioned near-wall cells in the vicinity of curved surfaces of bodies a global mesh deformation technique is used. The whole procedure is applied to a complex turbulent flow around a high-lift multiple element airfoil in take-off configuration using the Spalart–Allmaras turbulence model. The adaptation is controlled by as many indicators as there are equations involved in the problem. It is demonstrated that the proposed methodology performs rigorous local adaptive mesh refinement and automatically achieves grid independent results. Thus, interesting gains are obtained in terms of CPU time, memory requirement and user effort compared to single mesh computations. [Copyright &y& Elsevier]

The prediction of flow path of exhaust plume from the ship funnels is extremely complicated since the phenomenon is affected by a large number of parameters like wind velocity and direction, level of turbulence, geometry of the structures on ship’s deck, efflux velocity of smoke etc. To complicate the matters, the entire turbulent flow field is subject to abrupt changes as the yaw angle changes. In order to understand how the smoke is brought down to ship’s deck, it is necessary to have a knowledge of the funnel exhaust behavior very early in the design spiral of the ship by undertaking parametric investigation of the interaction effect between exhaust smoke and the ship superstructure. This paper presents such a parametric investigation on representative topside configurations of a generic frigate using computational fluid dynamics (CFD). The results presented have been analysed for a total of 112 different cases by varying velocity ratios and onset wind direction for four superstructure configurations. Use of both experimental and computational approaches has been made so that they become complementary to each other. The CFD simulation has been done using the computational code FLUENT version 6.0. Closure was achieved by using the standard k–ε turbulence model. The parametric study has demonstrated that CFD is a powerful tool to study the problem of exhaust smoke–superstructure interaction on ships and is capable of providing a means of visualising the path of the exhaust under different operating conditions very early in the design spiral of a ship. [Copyright &y& Elsevier]

Abstract: An efficient numerical method for the direct simulation of homogeneous turbulent flow has been obtained by combining a spectral forcing algorithm for homogeneous turbulence with a lattice-Boltzmann scheme for solution of the continuity and Navier–Stokes equations. The spectral forcing scheme of Alvelius [Alvelius K. Random forcing of three-dimensional homogeneous turbulence. Phys Fluids 1999;11(7):1880–89] is used which allows control of the power input by eliminating the force–velocity correlation in the Fourier domain and enables anisotropic forcing. A priori chosen properties such as the Kolmogorov length scale, the integral length scale and the integral time scale are recovered. This demonstrates that the scheme works accurately with the lattice-Boltzmann method and that all specific features of the forcing scheme are recovered in the lattice-Boltzmann implementation. [Copyright &y& Elsevier]

Abstract: The results of large eddy simulations (LES) of turbulent bubbly wake flows are presented. The LES technique was applied together with the Lagrangian particle dynamics method and a random flow generation (RFG) technique to the cases of a two-phase bubbly mixing layer and the high-Reynolds number bubbly ship-wake flows. The validation was performed on the experimental data for the bubbly mixing layer. Instantaneous distributions and probability density functions of bubbles in the wake were obtained using a joint LES/RFG approach. Separate estimates of bubble decay due to dissolution and buoyancy effects were obtained. The analysis of bubble agglomeration effects was done on the basis of experimental data for a turbulent vortex to satisfy one-way coupling that is used in this study. [Copyright &y& Elsevier]

• Bifurcation and Transition to Turbulence in Von Karman flow. • entropy viscosity Large Eddy Simulation model. • penalty method for moving solid obstacle in computational fluid dynamics. • Comparisons numerical simulations and laboratory experiments. Considering the current advances in experimental capabilities in fluid mechanics and the advances in computing power and numerical methods in computational fluid mechanics, a question that naturally arises is whether the two sets of techniques are approaching a level of sophistication sufficiently high to deliver results on turbulent flows in realistic geometries that are comparable. The purpose of this paper is to give elements of answers to this question by considering the so-called von Kármán flow where the fluid in a cylindrical container is driven by two counter-rotating impellers. We compare in the mentioned flow the torque and the flow topology obtained by experiments, direct numerical simulations (DNS), and large eddy simulations (LES) at various Reynolds numbers ranging from R e = O (10 2) to R e = O (10 5). In addition to validating the proposed LES model, the level of agreement that is observed between the numerical and the experimental data shows that the degree of accuracy of each of these techniques is reaching a threshold beyond which it is possible to use each of them with high confidence to explore and better understand turbulence in complex flows at R e = O (10 5) and beyond. [ABSTRACT FROM AUTHOR]

Abstract: The Euromech Colloquium 467 on Turbulent Flow and Noise Generation was held from July 18th to 20th, 2005 at the Centre International des Rencontres Mathématiques (CIRM) in Marseille, France. This international conference was attended by 65 participants coming from 10 different countries; 39 presentations were given, including three invited lectures by renowned experts: Fang Q. Hu (Old Dominion University, USA), Sanjiva K. Lele (Stanford University, USA) and Philippe Spalart (Boeing Commercial Airplanes, USA). The colloquium was followed in the same place by a summer school entitled Computational Aeroacoustics and CFD in Turbulent Flows, organized under the auspices of the Centre d’Eté de Mathématique et de Recherche Avancée en Calcul Scientifique (CEMRACS 2005) from July 18th to August 26th, 2005, which attracted around one hundred participants. This was the largest number of participants in this series of CEMRACS summer schools since its foundation in 1996 by Yvon Maday. The success of these two coupled events is a clear indication of the vitality of the subject of noise generation by flows and of its appeal to a wide population of scientists from various origins: acousticians, specialists of fluid mechanics and turbulence, applied mathematicians and specialists of numerical analysis. The Euromech Colloquium 467 was organized in the frame of the French–German research group FOR 508 Noise Generation in Turbulent Flows, whose aim is to develop and to assess novel noise simulation concepts for turbulent flow. A special focus is on the noise generated by turbulent jets. The research group FOR 508 is composed of seven projects, each of which is executed in joint cooperation of a German and a French research institution. The participants on the French side are: Ecole Polytechnique, Ecole Normale Supérieure, LMFA-Ecole Centrale de Lyon, IMFT de Toulouse, Université de Poitiers, Université de Provence (Aix-Marseille I) and Université Louis Pasteur Strasbourg. The German members of the research group FOR 508 are from the following institutions: Universität Stuttgart, RWTH Aachen, Technische Universitäten München, Berlin and Darmstadt. The research group started its work in January 2002 with financial support from the German Research Foundation Deutsche Forschungsgemeinschaft (DFG) and the French Centre National de la Recherche Scientifique (CNRS). [Copyright &y& Elsevier]