Large Eddy Simulation of the Flow Over a Three-Dimensional Hill

This paper investigates the use of LES for a flow around a three-dimensional axisymmetric hill. Two aspects of this simulation in particular are discussed here, the resolution and the inlet boundary conditions. In contrast to the LES of flows with sharp edge separations which do not require the near-wall dynamics to be fully resolved, the hill flow LES relies on the resolution of the upstream boundary layer in order to provoke the separation at a correct position. Although around 15 ×10⁶ computational cells were used, the resolution of streaky structures in the near-wall region that are important for a LES is not achieved. Two different inlet boundary conditions were used: the steady experimental profile and the time-dependent boundary conditions produced from DNS results of low Reynolds number channel flow. No significant improvement in the results was obtained with the unsteady inlet condition. This indicates that, although the unsteady inlet boundary conditions may be necessary for a successful LES of this flow, they must be followed with the resolution of the boundary layer for a successful LES.     

A Rapid and Accurate Switch from RANS to LES in Boundary Layers Using an Overlap Region

An efficient recycling algorithm is developed for injecting resolved turbulent content in a boundary layer as it switches from a Reynolds Averaged Navier-Stokes (RANS) type treatment to a Large Eddy Simulation (LES) type treatment inside a generalized Detached-Eddy Simulation (DES). The motivation is to use RANS in the thinnest boundary-layer area, following the original argument in favour of DES, and LES in the thicker boundary-layer areas especially approaching separation, to improve accuracy and possibly obtain unsteady outputs. The algorithm relies on an overlap of the RANS and LES domains and, therefore, the availability of both RANS and LES solutions in the recycling region, which is about 5 boundary-layer thicknesses long. This permits a smooth transfer of the turbulent stresses from this section to the LES inflow. The continuity of the skin-friction distribution is very good, reflecting the excellent viability of the resolved turbulence. The approach is validated in a flat-plate boundary layer and an airfoil near stall, with mild pressure gradient near the interface, and then applied to the compressible flow over an idealized airliner windshield wiper. The pressure fluctuations at reattachment are 12dB more intense than under a simple boundary layer at the same speed, and the output contains all the quantities needed to calculate the transmission of sound through the glass.     

A Zonal RANS/LES Approach for Noise Sources Prediction

Hybrid CFD/CAA methods have generally to be used for the numerical simulation of trailing-edge noise (see [9, 20] for instance). This study focuses on the first step of such hybrid methods, which is to predict the unsteady aerodynamic sources by the mean of a 3D unsteady simulation of the flow. Such a simulation is however generally still away from the numerical capabilities of ‘usual’ supercomputers. This paper investigates the use of a zonal LES method (based on the NLDE – Non-Linear Disturbance Equations – technique) for the numerical prediction of the aerodynamic noise sources. This method makes it possible to perform only zonal LES close to the main elements responsible of sound generation, while the overall configuration is only treated by a RANS approach. Attention will be paid to the specific boundary treatment at the interface between the RANS and LES regions. More precisely, the problem of the generation of turbulent inflow conditions for the LES region will be carefully addressed. The method is first assessed in the simulation of a flat plate ended by a blunted trailing-edge, and then applied to the simulation of the flow over a NACA0012 airfoil with blunted trailing-edge.     

High-order Large Eddy Simulations of Confined Rotor-Stator Flows

In many engineering and industrial applications, the investigation of rotating turbulent flow is of great interest. In rotor-stator cavities, the centrifugal and Coriolis forces have a strong influence on the turbulence by producing a secondary flow in the meridian plane composed of two thin boundary layers along the disks separated by a non-viscous geostrophic core. Most numerical simulations have been performed using RANS and URANS modelling, and very few investigations have been performed using LES. This paper reports on quantitative comparisons of two high-order LES methods to predict a turbulent rotor-stator flow at the rotational Reynolds number Re(=?Ωb ²/ν)?=4 × 10⁵. The classical dynamic Smagorinsky model for the subgrid-scale stress (Germano et al., Phys Fluids A 3(7):1760–1765, 1991) is compared to a spectral vanishing viscosity technique (Séverac & Serre, J Comp Phys 226(2):1234–1255, 2007). Numerical results include both instantaneous data and post-processed statistics. The results show that both LES methods are able to accurately describe the unsteady flow structures and to satisfactorily predict mean velocities as well as Reynolds stress tensor components. A slight advantage is given to the spectral SVV approach in terms of accuracy and CPU cost. The strong improvements obtained in the present results with respect to RANS results confirm that LES is the appropriate level of modelling for flows in which fully turbulent and transition regimes are involved.     

A Conceptual Study of Cavity Aeroacoustics Control Using Porous Media Inserts

In this study, an integrated flow simulation and aeroacoustics prediction methodology is applied to testing a sound control technique using porous inserts in an open cavity. Large eddy simulation (LES) combined with a three-dimensional Ffowcs Williams–Hawkings (FW–H) acoustic analogy is employed to predict the flow field, the acoustic sources and the sound radiation. The Darcy pressure – velocity law is applied to conceptually mimic the effect of porous media placed on the cavity floor and/or rear wall. Consequently, flow in the cavity could locally move in or out through these porous walls, depending on the local pressure differences. LES with “standard” subgrid-scale models for compressible flow is carried out to simulate the flow field covering the sound source and near fields, and the fully three-dimensional FW–H acoustic analogy is used to predict the sound field. The numerical results show that applying the conceptual porous media on cavity floor and/or rear wall could decrease the pressure fluctuations in the cavity and the sound pressure level in the far field. The amplitudes of the dominant oscillations (Rossiter modes) are suppressed and their frequencies are slightly modified. The dominant sound source is the transverse dipole term, which is significantly reduced due to the porous walls. As a result, the sound pressure in the far field is also suppressed. The preliminary study reveals that using porous-inserts is a promising technology for flow and sound radiation control.     

Hybrid methods for airframe noise numerical prediction

This paper describes some significant steps made towards the numerical simulation of the noise radiated by the high-lift devices of a plane. Since the full numerical simulation of such configuration is still out of reach for present supercomputers, some hybrid strategies have been developed to reduce the overall cost of such simulations. The proposed strategy relies on the coupling of an unsteady nearfield CFD with an acoustic propagation solver based on the resolution of the Euler equations for midfield propagation in an inhomogeneous field, and the use of an integral solver for farfield acoustic predictions.In the first part of this paper, this CFD/CAA coupling strategy is presented. In particular, the numerical method used in the propagation solver is detailed, and two applications of this coupling method to the numerical prediction of the aerodynamic noise of an airfoil are presented.Then, a hybrid RANS/LES method is proposed in order to perform some unsteady simulations of complex noise sources. This method allows for significant reduction of the cost of such a simulation by considerably reducing the extent of the LES zone. This method is described and some results of the numerical simulation of the three-dimensional unsteady flow in the slat cove of a high-lift profile are presented. While these results remain very difficult to validate with experiments on similar configurations, they represent up to now the first 3D computations of this kind of flow.     

Numerical prediction of turbulent boundary layer noise from a sharp-edged flat plate

An efficient hybrid uncorrelated wall plane waves–boundary element method (UWPW-BEM) technique is proposed to predict the flow-induced noise from a structure in low Mach number turbulent flow. Reynolds-averaged Navier-Stokes equations are used to estimate the turbulent boundary layer parameters such as convective velocity, boundary layer thickness, and wall shear stress over the surface of the structure. The spectrum of the wall pressure fluctuations is evaluated from the turbulent boundary layer parameters and by using semi-empirical models from literature. The wall pressure field underneath the turbulent boundary layer is synthesized by realizations of uncorrelated wall plane waves (UWPW). An acoustic BEM solver is then employed to compute the acoustic pressure scattered by the structure from the synthesized wall pressure field. Finally, the acoustic response of the structure in turbulent flow is obtained as an ensemble average of the acoustic pressures due to all realizations of uncorrelated plane waves. To demonstrate the hybrid UWPW-BEM approach, the self-noise generated by a flat plate in turbulent flow with Reynolds number based on chord Re_c = 4.9 × 10⁵ is predicted. The results are compared with those obtained from a large eddy simulation (LES)-BEM technique as well as with experimental data from literature.     

Detached Eddy Simulation of Flow around A-Airfoil

In this study the flow around the Aérospatiale A-airfoil at maximum lift (α = 13.3°) and Re = 2 × 10⁶ is investigated by RANS and DES to quantify the influence of transient flow patterns on the quality of the flow prediction. This flow features a highly unsteady pressure-induced trailing-edge separation. The fairly high Reynolds number makes LES rather unattractive from a practical point of view as the numerical costs scale with ≈ Re^1.95 [9] and lead to an unreasonable effort to carry out these simulations. Like LES, DES is designed to capture the unsteady flow features and consequently is supposed to give results superior to RANS. The RANS solution captures the separation, but is notable to predict transient vortex shedding. The application of DES to a 2D domain leads to results comparable to those obtained by 2D-RANS. Only the 3D-DES is capable of predicting both the unsteady flow features and the development of the boundary layer on the airfoil. As expected, the spanwise resolution has a strong impact on the vortex shedding and on the size of the separation bubble. Attention must be paid to both accommodating the full extent of the flow structures and providing the adequate spatial resolution in order to accomplish satisfactory results – that compare favourably with LES – at only a fraction of computational cost.     

Time-resolved stereo PIV measurements of shock–boundary layer interaction on a supercritical airfoil

Time-resolved stereo particle-image velocimetry (TR-SPIV) and unsteady pressure measurements are used to analyze the unsteady flow over a supercritical DRA-2303 airfoil in transonic flow. The dynamic shock wave–boundary layer interaction is one of the most essential features of this unsteady flow causing a distinct oscillation of the flow field. Results from wind-tunnel experiments with a variation of the freestream Mach number at Reynolds numbers ranging from 2.55 to 2.79 × 10⁶ are analyzed regarding the origin and nature of the unsteady shock–boundary layer interaction. Therefore, the TR-SPIV results are analyzed for three buffet flows. One flow exhibits a sinusoidal streamwise oscillation of the shock wave only due to an acoustic feedback loop formed by the shock wave and the trailing-edge noise. The other two buffet flows have been intentionally influenced by an artificial acoustic source installed downstream of the test section to investigate the behavior of the interaction to upstream-propagating disturbances generated by a defined source of noise. The results show that such upstream-propagating disturbances could be identified to be responsible for the upstream displacement of the shock wave and that the feedback loop is formed by a pulsating separation of the boundary layer dependent on the shock position and the sound pressure level at the shock position. Thereby, the pulsation of the separation could be determined to be a reaction to the shock motion and not vice versa.     

Motion of a vertical wall fixed on springs under the action of surface waves

A two-dimensional unsteady hydroelastic problem of interaction between surface waves and a moving vertical wall fixed on springs is studied. An analytical solution of the problem is constructed using a linear approximation, and a numerical solution within the framework of a nonlinear model of a potential fluid flow is found by a complex boundary element method. By means of analysis of the linear and nonlinear solutions, it is found that the linear solution can be used to predict some important characteristics of the wall motion and the fluid flow in the case of moderate wave amplitudes.     

Noise Investigation of a High Subsonic, Moderate Reynolds Number Jet Using a Compressible Large Eddy Simulation

This study investigates the noise radiated by a subsonic circular jet with a Mach number of 0.9 and a Reynolds number of 65000 computed by a compressible Large Eddy Simulation (LES). First, it demonstrates the feasibility of using LES to predict accurately both the flow field and the sound radiation on a domain including the acoustic field. Mean flow parameters, turbulence intensities, velocity spectra and integral length scales are in very good agreement with experimental data. The noise generated by the jet, provided directly by the simulation, is also consistent with measurements in terms of sound pressure spectra, levels and directivity. The apparent location of the sound sources is at the end of the potential core in accordance with some experimental observations at similar Reynolds numbers and Mach numbers. Second, the noise generation mechanisms are discussed in an attempt to connect the flow field with the acoustic field. This study shows that for the simulated moderate Reynolds number jet, the predominant sound radiation in the downstream direction is associated with the breakdown of the shear layers in the central jet zone. Received 24 January 2002 and accepted 16 July 2002 Published online 3 December 2002 RID="*" ID="*" A preliminary version of some of the results presented here was reported in AIAA Paper 2000–2009 presented at the 6th AIAA/CEAS Aeroacoustics Conference in Lahaina, Hawaii, June 2000. Computing time was supplied by the Institut du Développement et des Ressources en Informatique Scientifique (IDRIS – CNRS). Communicated by T.B. Gatski     

Buffeting in transonic flow prediction using time‐dependent turbulence model

In transonic flow conditions, the shock wave/turbulent boundary layer interaction and flow separations on wing upper surface induce flow instabilities, ‘buffet’, and then the buffeting (structure vibrations). This phenomenon can greatly influence the aerodynamic performance. These flow excitations are self‐sustained and lead to a surface effort due to pressure fluctuations. They can produce enough energy to excite the structure. The objective of the present work is to predict this unsteady phenomenon correctly by using unsteady Navier–Stokes‐averaged equations with a time‐dependent turbulence model based on the suitable (k–ε) turbulent eddy viscosity model. The model used is based on the turbulent viscosity concept where the turbulent viscosity coefficient (Cμ) is related to local deformation and rotation rates. To validate this model, flow over a flat plate at Mach number of 0.6 is first computed, then the flow around a NACA0012 airfoil. The comparison with the analytical and experimental results shows a good agreement. The ONERA OAT15A transonic airfoil was chosen to describe buffeting phenomena. Numerical simulations are done by using a Navier–Stokes SUPG (streamline upwind Petrov–Galerkin) finite‐element solver. Computational results show the ability of the present model to predict physical phenomena of the flow oscillations. The unsteady shock wave/boundary layer interaction is described. Copyright © 2005 John Wiley & Sons, Ltd.     

Prediction of periodic boundary layers

A relatively simple, yet efficient and accurate finite difference method is developed for the solution of the unsteady boundary layer equations for both laminar and turbulent flows. The numerical procedure is subjected to rigorous validation tests in the laminar case, comparing its predictions with exact analytical solutions, asymptotic solutions, and/or experimental results. Calculations of periodic laminar boundary layers are performed from low to very high oscillation frequencies, for small and large amplitudes, for zero as well as adverse time-mean pressure gradients, and even in the presence of significant flow reversal. The numerical method is then applied to predict a relatively simple experimental periodic turbulent boundary layer, using two well-known quasi-steady closure models. The predictions are shown to be in good agreement with the measurements, thereby demonstrating the suitability of the present numerical scheme for handling periodic turbulent boundary layers. The method is thus a useful tool for the further development of turbulence models for more complex unsteady flows.     

The transmission of turbulent boundary layer unsteady pressure and shear stress through a viscoelastic layer

The transmission of unsteady pressure and shear stress, generated by a turbulent boundary layer in water, through a viscoelastic layer backed by a rigid plate is investigated. Analytical models are used to estimate the unsteady pressure and shear stress from 10 to 1000 Hz for a flat plate boundary layer with zero pressure gradient. Additionally, models for the transfer of the unsteady pressures and shear stress through the viscoelastic layer are developed. The models are used to predict the unsteady pressure fluctuations, or flow noise, which would be seen by a finite size sensor embedded under the elastomer layer. The unsteady pressure levels are found to be 20 dB greater than the unsteady shear stress levels across all frequency ranges computed, in agreement with recent measurements. The unsteady pressure transfer functions have a peak at the shear wavenumber and are larger than the shear stress transfer magnitudes from 10 to 50 Hz. The unsteady shear stress transfer functions have a peak at the acoustic wavenumber and are larger than the pressure transfer magnitudes from 50 to 1000 Hz. Over the frequency range examined, the unsteady pressures were found to be the dominant contributor to the sensor flow noise due to the considerably larger magnitude of the unsteady pressures on the top of the viscoelastic layer.     

Highly Resolved LES and URANS of Turbulent Buoyancy-Driven Flow Within Inclined Differentially-Heated Enclosures

In the present study, both LES and unsteady RANS computations are presented, for turbulent natural convection of air inside differentially-heated rectangular tilted cavities using a finite volume code (Code_Saturne), for an aspect ratio of H/L?=?28.6 and Rayleigh number of 0.86×10⁶. Attention is focused on two angles of inclination: 15° to the horizontal with hot lower and cold upper wall, the 15° unstable case, and the mirror image of this case where the angle is the same but with a hot upper and cold lower wall, the 15° stable case. In accordance with recent experimental data, the LES computations for both the stable and unstable tilted cavities returned three-dimensional time-averaged flow fields. In the case of the unstably stratified enclosure, the flow is highly unsteady with coherent turbulent structures in the core of the enclosure. Time-averaged temperature, velocity and resolved turbulence intensities resulting from LES computations show close agreement to measured data. Subsequent comparisons of different URANS schemes with LES are used in order to explore to what extent these models are able to reproduce the large-scale unsteady flow structures. All URANS schemes have been found to be able to reproduce the 3-D unsteady flow features present in the 15° unstable cavity. However, the low-Reynolds-number model tested, as well as requiring a high resolution near-wall grid, also needed a finer grid in the core region than the high-Reynolds-number models, thus making it computationally very expensive. Flow within the 15° stable cavity also shows some 3-D features, although it is significantly less unsteady, and the URANS models tested here have been less successful in reproducing this flow pattern. The overall heat transfer is presented here for both differentially heated enclosures.     

风口区空气流动的有限元模拟与实验研究

通风和空调出风口的射流特性主要取决于出风口截面之前的一段管道内的流动状况,研究送风口的射流偏转特性有重要实用价值.应用湍流大涡模拟技术结合Taylor-Galerkin有限单元法对工程中常见工况下的风口区管道内的三维流场进行了数值模拟和实验研究,数值计算结果与实验结果符合较好.表明湍流大涡模拟方法适合于边界形状复杂,存在各向异性的大尺度涡的内流情况,能可靠地预报风口区空气流动的射流偏转特性.     

Simulation of supersonic turbulent flow in the vicinity of an inclined backward-facing step

Large eddy simulation (LES) is a viable and powerful tool to analyse unsteady three-dimensional turbulent flows. In this article, the method of LES is used to compute a plane turbulent supersonic boundary layer subjected to different pressure gradients. The pressure gradients are generated by allowing the flow to pass in the vicinity of an expansion–compression ramp (inclined backward-facing step with leeward-face angle of 25°) for an upstream Mach number of 2.9. The inflow boundary condition is the main problem for all turbulent wall-bounded flows. An approach to solve this problem is to extract instantaneous velocities, temperature and density data from an auxiliary simulation (inflow generator). To generate an appropriate realistic inflow condition to the inflow generator itself the rescaling technique for compressible flows is used. In this method, Morkovin's hypothesis, in which the total temperature fluctuations are neglected compared with the static temperature fluctuations, is applied to rescale and generate the temperature profile at inlet. This technique was successfully developed and applied by the present author for an LES of subsonic three-dimensional boundary layer of a smooth curved ramp. The present LES results are compared with the available experimental data as well as numerical data. The positive impact of the rescaling formulation of the temperature is proven by the convincing agreement of the obtained results with the experimental data compared with published numerical work and sheds light on the quality of the developed compressible inflow generator.     

基于奇异分解和自适应BEM积分算法的水下航行体机械噪声预报

提出一种改进的声学边界元法（M-BEM）用于准确计算水下航行体发动机振动引起的近场辐射噪声。分别采用奇异分解技术和自适应边界元积分算法解决了Helmholtz积分方程在求解近场声压时出现的超奇异积分和奇异积分问题。采用一脉动球源的声辐射算例对方法进行验证,数值解与精确解误差小于1．5dB。结合有限元方法并考虑流固耦合作...     

Large Eddy Simulation and PIV Measurements of Unsteady Premixed Flames Accelerated by Obstacles

In gas explosions, the unsteady coupling of the propagating flame and the flow field induced by the presence of blockages along the flame path produces vortices of different scales ahead of the flame front. The resulting flame–vortex interaction intensifies the rate of flame propagation and the pressure rise. In this paper, a joint numerical and experimental study of unsteady premixed flame propagation around three sequential obstacles in a small-scale vented explosion chamber is presented. The modeling work is carried out utilizing large eddy simulation (LES). In the experimental work, previous results (Patel et al., Proc Combust Inst 29:1849–1854, 2002) are extended to include simultaneous flame and particle image velocimetry (PIV) measurements of the flow field within the wake of each obstacle. Comparisons between LES predictions and experimental data show a satisfactory agreement in terms of shape of the propagating flame, flame arrival times, spatial profile of the flame speed, pressure time history, and velocity vector fields. Computations through the validated model are also performed to evaluate the effects of both large-scale and sub-grid scale (SGS) vortices on the flame propagation. The results obtained demonstrate that the large vortical structures dictate the evolution of the flame in qualitative terms (shape and structure of the flame, succession of the combustion regimes along the path, acceleration-deceleration step around each obstacle, and pressure time trend). Conversely, the SGS vortices do not affect the qualitative trends. However, it is essential to model their effects on the combustion rate to achieve quantitative predictions for the flame speed and the pressure peak.     

Computational analysis and flow structure of a transitional separated-reattached flow over a surface mounted obstacle and a forward-facing step

Large-eddy simulation (LES) of transitional separating-reattaching flow on a two-dimensional square surface mounted obstacle and a forward facing step has been performed using a dynamic sub-grid scale model. The Reynolds number based on the uniform inlet velocity and the obstacle/step height is 4.5 × 10³. The mean LES results for both the obstacle and step flow compare reasonably well with the available experimental and DNS data.

The flow structures upstream of the surface-mounted obstacle (referred to hereafter as obstacle) and the forward-facing step (referred to hereafter as FFS) consist of unstable two-dimensional structures and coherent rib-shaped structures. These structures with the aid of 3D streamline visualisation strongly indicate that the upstream separation bubble is a closed one rather than an open one in the sense that there is little evidence to suggest that there is fluid injection from the upstream separation region into the downstream separated region for the two geometries. The spectra and time history for the velocities and pressure fields at locations immediately upstream of the obstacle and FFS (including the recirculation region) were analysed using both the Fourier and wavelet transforms and revealed the unsteady nature of the recirculation region upstream of the obstacle and FFS.

The transition process has been elucidated using both 2D and 3D flow visualisation of the flow. In both geometries (obstacle and FFS), the separated boundary layer downstream of the leading edge shows 2D nature and roll-up shortly downstream of the separation line leading to 2D K-H rolls to be shed from the leading edge. Coherent structures such as the λ-shaped and rib-like vortices commonly associated with a flat plate boundary layer and also found in the separated-reattached flow of a blunt leading edge plate aligned horizontally to a flow are not common in the separated-reattached flow over the obstacle and FFS.