Showing papers in "Theoretical and Computational Fluid Dynamics in 2005"

A rod-airfoil experiment as a benchmark for broadband noise modeling

Abstract: A low Mach number rod-airfoil experiment is shown to be a good benchmark for numerical and theoretical broadband noise modeling. The benchmarking approach is applied to a sound computation from a 2D unsteady-Reynolds-averaged Navier–Stokes (U-RANS) flow field, where 3D effects are partially compensated for by a spanwise statistical model and by a 3D large eddy simulation. The experiment was conducted in the large anechoic wind tunnel of the Ecole Centrale de Lyon. Measurements taken included particle image velocity (PIV) around the airfoil, single hot wire, wall pressure coherence, and far field pressure. These measurements highlight the strong 3D effects responsible for spectral broadening around the rod vortex shedding frequency in the subcritical regime, and the dominance of the noise generated around the airfoil leading edge. The benchmarking approach is illustrated by two examples: In both cases, the ability of computational fluid dynamics to model the source mechanisms and of the CAA approach to predict the far field are assessed separately.

Towards a new partially integrated transport model for coarse grid and unsteady turbulent flow simulations

Abstract: A new two-equation model is proposed for large eddy simulations (LESs) using coarse grids. The modeled transport equations are obtained from a direct transposition of well-known statistical models by using multiscale spectrum splitting given by the filtering operation applied to the Navier–Stokes equations. The model formulation is compatible with the two extreme limits that are on one hand a direct numerical simulation and on the other hand a full statistical modeling. The characteristic length scale of subgrid turbulence is no longer given by the spatial discretization step size, but by the use of a dissipation equation. The proposed method is applied to a transposition of the well-known k-e statistical model, but the same method can be developed for more advanced closures. This approach is intended to contribute to non-zonal hybrid models that bridge Reynolds-averaged Navier–Stokes (RANS) and LES, by using a continuous change rather than matching zones. The main novelty in the model is the derivation of a new e equation for LES that is formally consistent with RANS when the filter width is very large. This approach is dedicated to applications to non-equilibrium turbulence and coarse grid simulations. An illustration is made of large eddy simulations of turbulence submitted to periodic forcing. The model is also an alternative approach to hybrid models.

Hybrid methods for airframe noise numerical prediction

Abstract: 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.

Vortex ring modelling of toroidal bubbles

Abstract: During the collapse of a bubble near a surface, a high-speed liquid jet often forms and subsequently impacts upon the opposite bubble surface. The jet impact transforms the originally singly-connected bubble to a toroidal bubble, and generates circulation in the flow around it. A toroidal bubble simulation is presented by introducing a vortex ring seeded inside the bubble torus to account for the circulation. The velocity potential is then decomposed into the potential of the vortex ring and a remnant potential. Because the remnant potential is continuous and satisfies the Laplace equation, it can be modelled by the boundary-integral method, and this circumvents an explicit domain cut and associated numerical treatment. The method is applied to study the collapse of gas bubbles in the vicinity of a rigid wall. Good agreement is found with the results of Best (J. Fluid Mech. 251 79–107, 1993), obtained by a domain cut method. Examination of the pressure impulse on the wall during jet impact indicates that the high-speed liquid jet has a significant potential for causing damage to a surface. There appears to be an optimal initial distance where the liquid jet is most damaging.

A non-oscillatory balanced scheme for an idealized tropical climate model

Abstract: We propose a non-oscillatory balanced numerical scheme for a simplified tropical climate model with a crude vertical resolution, reduced to the barotropic and the first baroclinic modes. The two modes exchange energy through highly nonlinear interaction terms. We consider a periodic channel domain, oriented zonally and centered around the equator and adopt a fractional stepping–splitting strategy, for the governing system of equations, dividing it into three natural pieces which independently preserve energy. We obtain a scheme which preserves geostrophic steady states with minimal ad hoc dissipation by using state of the art numerical methods for each piece: The f-wave algorithm for conservation laws with varying flux functions and source terms of Bale et al. (2002) for the advected baroclinic waves and the Riemann solver-free non-oscillatory central scheme of Levy and Tadmor (1997) for the barotropic-dispersive waves. Unlike the traditional use of a time splitting procedure for conservation laws with source terms (here, the Coriolis forces), the class of balanced schemes to which the f-wave algorithm belongs are able to preserve exactly, to the machine precision, hydrostatic (geostrophic) numerical-steady states. The interaction terms are gathered into a single second order accurate predictor-corrector scheme to minimize energy leakage. Validation tests utilizing known exact solutions consisting of baroclinic Kelvin, Yanai, and equatorial Rossby waves and barotropic Rossby wave packets are given.

Generation of compressible modes in MHD turbulence

Abstract: Astrophysical turbulence is magnetohydrodynamic (MHD) in nature. We discuss fundamental properties of MHD turbulence and in particular the generation of compressible MHD waves by Alfvenic turbulence and show that this process is inefficient. This allows us to study the evolution of different types of MHD perturbations separately. We describe how to separate MHD fluctuations into three distinct families: Alfven, slow, and fast modes. We find that the degree of suppression of slow and fast modes production by Alfvenic turbulence depends on the strength of the mean field. We review the scaling relations of the modes in strong MHD turbulence. We show that Alfven modes in compressible regime exhibit scalings and anisotropy similar to those in incompressible regime. Slow modes passively mimic Alfven modes. However, fast modes exhibit isotropy and a scaling similar to that of acoustic turbulence both in high and low β plasmas. We show that our findings entail important consequences for star formation theories, cosmic ray propagation, dust dynamics, and gamma ray bursts. We anticipate many more applications of the new insight to MHD turbulence and expect more revisions of the existing paradigms of astrophysical processes as the field matures.

Investigation of the onset of thermo-bioconvection in a suspension of oxytactic microorganisms in a shallow fluid layer heated from below

Abstract: This paper investigates the combined effect of density stratification due to oxytactic upswimming and heating from below on the stability of a suspension of motile oxytactic microorganisms in a shallow fluid layer. Different from traditional bioconvection, thermo-bioconvection has two destabilizing mechanisms that contribute to creating the unstable density stratification. This problem may be relevant to a number of geophysical applications, such as the investigation of the dynamics of some species of thermophiles (heat loving microorganisms) living in hot springs. By performing a linear stability analysis, we obtained a correlation between the critical value of the bioconvection Rayleigh number and the traditional, “thermal” Rayleigh number. It is established that heating from below makes the system more unstable and helps the development of bioconvection.

A non-oscillatory balanced scheme for an idealized tropical climate model Part II: Nonlinear coupling and moisture effects

Abstract: We use the non-oscillatory balanced numerical scheme developed in Part I to track the dynamics of a dry highly nonlinear barotropic/baroclinic coupled solitary wave, as introduced by Biello and Majda (2004), and of the moisture fronts of Frierson et al. (2004) in the presence of dry gravity waves, a barotropic trade wind, and the beta effect. It is demonstrated that, for the barotropic/baroclinic solitary wave, except for a little numerical dissipation, the scheme utilized here preserves total energy despite the strong interactions and exchange of energy between the baroclinic and barotropic components of the flow. After a short transient period where the numerical solution stays close to the asymptotic predictions, the flow develops small scale eddies and ultimately becomes highly turbulent. It is found here that the interaction of a dry gravity wave with a moisture front can either result in a reflection of a fast moistening front or the pure extinction of the precipitation. The barotropic trade wind stretches the precipitation patches and increases the lifetime of the moisture fronts which decay naturally by the effects of dissipation through precipitation while the Coriolis effect makes the moving precipitation patches disappear and appear at other times and places.

Noise radiated by a non-isothermal, temporal mixing layer. Part II: Prediction using DNS in the framework of low Mach number approximation

Abstract: The ability of Lighthill's analogy to predict the sound radiated by a transitional mixing layer is evaluated by means of direct numerical simulation (DNS). The specific case of low Mach number flows with density variations is investigated. In order to limit the global computational cost, the acoustic source information is based on numerical results where the sound waves have been removed. It is shown that the low Mach number approximation coupled with the acoustic analogy can lead to very accurate predictions for the radiated sound if the acoustic sources in Lighthill's equation are taken into account carefully. Results for the acoustic intensity deduced from a repeated use of the Lighthill's analogy over a wide range of Mach numbers allow us to discuss the adequacy of scaling laws proposed by previous authors (J. Sound Vib. 28(3), 563–585, 1973; 31(4), 391–397, 1973; 48(1), 95–111, 1976) for the prediction of noise from hot jets.

Spatial stability of the incompressible corner flow

Abstract: The linear spatial stability of the incompressible corner flow under pressure gradient has been studied. A self-similar form has been used for the mean flow, which reduces the related problem to the solution of a two-dimensional problem. The stability problem was formulated using the parabolised stability equations (PSE) and results were obtained for the viscous modes at medium and high frequencies. The related N-factors indicate that the flow is stable at these frequencies, but probably unstable for small frequencies. Furthermore the inviscid mode for each mean flow was obtained and the results indicate that its importance increases considerably with an increase in the adverse pressure gradient. Finally the dependence of the stability characteristics on the extent of the domain is also considered.

Large Eddy Simulation of the sound field of a round turbulent jet

Abstract: In this paper we will use Large Eddy Simulation (LES) to obtain the flow field of a turbulent round jet at a Reynolds number based on the jet orifice velocity of 11000. In the simulations it is assumed that the flow field is incompressible. The acoustic field of the jet is calculated with help of the Lighthill acoustic analogy. The coupling between the flow solver and the acoustic solver is discussed in detail. The Mach number used in the acoustic calculation was equal to 0.6. It is shown that the decay of the jet centerline velocity and centerline rms are in good agreement with experimental data of [12]. Furthermore, it is shown that the influence of the LES modeling on the acoustic field is very small, if the dynamic subgrid model is used.

Free surface flow under gravity and surface tension due to an applied pressure distribution: I Bond number greater than one-third

Abstract: We consider steady free surface two-dimensional flow due to a localized applied pressure distribution under the effects of both gravity and surface tension in water of constant depth, and in the presence of a uniform stream. The fluid is assumed to be inviscid and incompressible, and the flow is irrotational. The behavior of the forced nonlinear waves is characterized by three parameters: the Froude number, F, the Bond number, τ > 1/3, and the magnitude and sign of the pressure forcing parameter ɛ. The fully nonlinear wave problem is solved numerically by using a boundary integral method. For small amplitude waves and F 1/3, there exists both elevation and depression waves. In some cases, a limiting configuration in the form of a trapped bubble occurs in the depression wave solutions.

An improved method for calculating flow past flapping and hovering airfoils

Abstract: A method is reported here for calculating unsteady aerodynamics of hovering and flapping airfoil for two-dimensional flow via the following improved methodologies: (a) a correct formulation of the problem using stream function (ψ) and vorticity (ω) as dependent variables; (b) calculating loads and moment by a new method to solve the governing pressure Poisson equation (PPE) in a truncated part of the computational domain on a nonstaggered grid; (c) accurate solution using high accuracy compact difference scheme for the vorticity transport equation (VTE) and (d) accelerating the computations by using a high-order filter after each time step of integration. These have been used to solve Navier–Stokes equation for flow past flapping and hovering NACA 0014 and 0015 airfoils at typical Reynolds numbers relevant to the study of unsteady aerodynamics of micro air vehicle (MAV) and insect/bird flight.

Magma flow through elastic-walled dikes

Abstract: A convection–diffusion model for the averaged flow of a viscous, incompressible magma through an elastic medium is considered. The magma flows through a dike from a magma reservoir to the Earth’s surface; only changes in dike width and velocity over large vertical length scales relative to the characteristic dike width are considered. The model emerges when nonlinear inertia terms in the momentum equation are neglected in a viscous, low-speed approximation of a magma flow model coupled to the elastic response of the rock. Stationary- and traveling-wave solutions are presented in which a Dirichlet condition is used at the magma chamber; and either a (i) free-boundary condition, (ii) Dirichlet condition, or (iii) choked-flow condition is used at the moving free or fixed-top boundary. A choked-flow boundary condition, generally used in the coupled elastic wave and magma flow model, is also used in the convection–diffusion model. The validity of this choked-flow condition is illustrated by comparing stationary flow solutions of the convection–diffusion and coupled elastic wave and magma flow model for parameter values estimated for the Tolbachik volcano region in Kamchatka, Russia. These free- and fixed-boundary solutions are subsequently explored in a conservative, local discontinuous Galerkin finite-element discretization. This method is advantageous for the accurate implementation of the choked flow and free-boundary conditions. It uses a mixed Eulerian–Lagrangian finite element with special infinite curvature basis function near the free boundary and ensures positivity of the mean aperture subject to a time-step restriction. We illustrate the model further by simulating magma flow through host rock of variable density, and magma flow that is quasi-periodic due to the growth and collapse of a lava dome.

Spatial characterization of the numerically simulated vorticity fields of a flow in a flume

Abstract: The topology of large scale structures in a turbulent boundary layer is investigated numerically. Spatial characteristics of the large scale structure are presented through an original method, proper orthogonal decomposition (POD) of the three-dimensional vorticity fields. The DNS results, obtained by Tiselj et al. [23] for a fully developed turbulent flow in a flume, are used in the present work to analyze coherent structures with the proposed methodology. In contrast to the reconstruction methods that use instantaneous flow quantities, this approach utilizes the whole dataset of the numerical simulation. The analysis uses one thousand 3D vorticity fields from 50000 time steps of the simulation for the Reynolds number of 2600 (the turbulent Reynolds number Re*=171). The computational domain is 2146×171×537 wall units and the grid resolution is 128×65×72 points (in streamwise, wall-normal and spanwise directions, respectively). Experimental results obtained by using particle image velocimetry (PIV) in a fully developed turbulent boundary layer in a flume, which were analyzed with the same statistical characterization method, are in agreement with the DNS analysis: the dominant vortical structure appears to have a longitudinal streamwise orientation, an inclination angle of about 8°, streamwise length of several hundred wall units, and a distance between the neighboring structures of about 100 wall units in the spanwise direction.

Diffusion velocity for a three-dimensional vorticity field

Abstract: A discussion is presented on the existence of a diffusion velocity for the vorticity vector that satisfies extensions of the Helmholtz vortex laws in a three-dimensional, incompressible, viscous fluid flow. A general form for the diffusion velocity is derived for a complex-lamellar vorticity field that satisfies the property that circulation is invariant about a region that is advected with the sum of the fluid velocity and the diffusion velocity. A consequence of this property is that vortex lines will be material lines with respect to this combined velocity field. The question of existence of diffusion velocity for a general three-dimensional vorticity field is shown to be equivalent to the question of existence of solutions of a certain Fredholm equation of the first kind. An example is given for which it is shown that a diffusion velocity satisfying this property does not, in general, exist. Properties of the simple expression for diffusion velocity for a complex-lamellar vorticity field are examined when applied to the more general case of an arbitrary three-dimensional flow. It is found that this form of diffusion velocity, while not satisfying the condition of circulation invariance, nevertheless has certain desirable properties for computation of viscous flows using Lagrangian vortex methods. The significance and structure of the noncomplex-lamellar part of the viscous diffusion term is examined for the special case of decaying homogeneous turbulence.

Sound and vorticity interactions: transmission and scattering

Abstract: We review several aspects of the propagation of sound in vortical flows. We restrict ourselves to isothermal, humidity-free flows at low Mach number M. Since vorticity plays a major role in vortex-flow interactions we focus on vortical flows. We consider two main canonical situations. The first concerns the transmission of sound. We analyze the evolution of acoustic wavefronts as they propagate across a single vortex. The second situation addresses the scattering of sound waves by nonstationary vortices. We study the evolution of the acoustic pressure emitted in the far field, at an angle with the initial direction of propagation. In this geometry one performs direct spectroscopy of the flow vorticity field. In each case, we review theoretical results and compare with experimental measurements and numerical simulations when available. We also briefly report how the following new acoustic techniques have recently been used to study complex or turbulent flows: time-resolved acoustic spectroscopy, speckle interferometry and Lagrangian particle tracking.

Reynolds number dependence of two-dimensional laminar co-rotating vortex merging

Abstract: The paper reports unsteady Navier–Stokes calculations of laminar two-dimensional co-rotating vortex merging for various Reynolds numbers. The unsteady, incompressible two-dimensional Navier–Stokes equations were solved with fourth-order Runge–Kutta temporal discretization and fourth-order symmetric compact schemes for spatial discretization. Calculations of the unsteady Taylor vortex benchmark showed that fourth-order accurate solutions for all primitive variables were indeed achieved. Calculations for a pair of equal-strength co-rotating vortices show good agreement with reported direct numerical simulation and experiments for the evolution of the separation distance and core radius. It is found that the time required for merging is inversely proportional to the square root of the Reynolds number. According to previous experimental research, it was also found that complete merging in laminar regime undergoes four stages with physical meaning. The physical mechanism responsible for the merging process is investigated and it is found that the antisymmetric vorticity dynamics plays an important role until full merging.

On the consequences of material frame-indifference in algebraic stress models

Abstract: The principle of material frame-indifference (MFI) is a fundamental and controversial principle of continuum mechanics that has been invoked recently to derive nonlinear algebraic models for stresses of viscoelastic liquids. The purpose of the present study is to identify regions of a flow field where MFI should be considered. Such regions are identified by computing the angular velocity of the principal directions of the rate-of-deformation tensor in order to obtain an Euclidean objective vorticity tensor. An analysis is carried out for uniform shear and extensional flows, and for a Couette flow. The method is then applied to the planar flow through an abrupt 4:1 contraction and to the two-dimensional stream past a circular cylinder. The main results are: (1) MFI should be taken into account in regions characterized by the transition between two different kinematics and a significant velocity magnitude, and (2) MFI can be safely ignored in regions of pure viscometric behaviour as well as in recirculation regions. The consequences of MFI being taken into account are then examined upon using the Euclidean objective vorticity tensor in a simple algebraic constitutive law for viscoelastic fluids.

Accurate estimate of disturbance amplitude variation from solution of minimal composite stability theory

Abstract: Minimal composite theory (Proc. R. Soc. Lond., Ser. A, 453 2537–2549, 1979) shows that, at the lowest order in the reciprocal of the local flow Reynolds number R, the stability of a spatially developing similarity flow may be described by an ordinary differential equation in the similarity coordinate. It is, in principle, not possible to determine the dependence of the disturbance amplitude on the streamwise coordinate solely from such an ordinary differential equation. However, noting that, to O(R−2/3), the dependence of the eigenfunction on the normal coordinate is identical in both the full non-parallel and minimal composite theories, and using a method due to Gaster, we show how the streamwise variation of disturbance amplitude can be determined to O(R−1 without solving a partial differential equation, although knowledge of the partial differential operator is required. Comparison with the DNS results of Fasel and Konzelmann shows excellent agreement with the present results. Furthermore, especially in strong adverse pressure gradients, the present amplitude ratio estimates are within 3% of the full non-parallel theory, whereas the Orr–Sommerfeld results show an underestimate by 26%.

Analysis of pipe flow transition.Part II. Energy transfer

Abstract: The present article describes the results from a study of nonlinear mechanisms at work during the process of transition to turbulence in pipe flows. Using an accurate hybrid finite-difference code for the simulation of unsteady incompressible pipe flow, we have performed a direct numerical simulation designed to model experiments performed by Han, Tumin and Wygnanski [12]. Based on these numerical data, we have conducted a meticulous investigation of the dynamic interactions of the structures and flow modes that can be observed during this process. Based on this study, we can paint a detailed picture of the dynamical interactions of flow structures during both the linear and nonlinear stages of pipe flow transition. While this picture does have some similarities to earlier proposed mechanisms, we find that even for the simple cases considered here the structure of the pertinent interactions is much richer than suggested by these earlier models.

Pressure force transition and mixture morphology evolution in heterogeneous mixtures

Abstract: The pressure gradient terms in multifluid dynamical descriptions with separate momentum equations for each species or material have different forms depending on the mixture morphology. As the mixture evolves from chunk to molecular mixing, the form of the pressure gradient terms should evolve accordingly. This paper presents a new formulation of the pressure gradient terms which exhibits the appropriate smooth transition in form as the mixture evolves from chunk to molecular mixing, including intermediate stages in which both types of mixing are present. This transition is accomplished by introducing appropriate mixture morphology variables βi to describe the degree of molecular vs. chunk mixing for each species. A simple provisional relaxation-time model for the time evolution of the βi is also presented.

Effect of pulsations on the stability of a gas column

Abstract: The effect of radial pulsations on the stability of a compressible cylindrical gas column surrounded by an ambient liquid is discussed. In the absence of pulsations, the stationary interface is susceptible to the Rayleigh capillary instability, which promotes the growth of longitudinal waves whose wave length is larger than 2π times the column radius, irrespective of the Reynolds number. A Floquet stability analysis for potential flow shows that the pulsations further destabilize the interface by extending the range of unstable wave numbers to a sequence of islands. A similar stability analysis for Stokes flow shows that the pulsations also have a destabilizing influence, though the presence of an insoluble surfactant has a competing stabilizing influence that may cause an overall reduction in the range of unstable wave numbers.

Direct numerical simulation of homogeneous stratified rotating turbulence

Abstract: The effects of the Prandtl number on stratified rotating turbulence have been studied in homogeneous turbulence by using direct numerical simulations and a rapid distortion theory. Fluctuations under strong stable-density stratification can be theoretically divided into the WAVE and the potential vorticity (PV) modes. In low-Prandtl-number fluids, the WAVE mode deteriorates, while the PV mode remains. Imposing rotation on a low-Prandtl-number fluid makes turbulence two-dimensional as well as geostrophic; it is found from the instantaneous turbulent structure that the vortices merge to form a few vertically-elongated vortex columns. During the period toward two-dimensionalization, the vertical vortices become asymmetric in the sense of rotation.

On the front condition for compositional gravity currents

Abstract: A two-dimensional potential flow is employed to derive the front condition of the gravity current. The derivation starts from the balance between the static pressure of the gravity current and the form drag imposed on the gravity current by the ambient fluid. After employing Bernoulli’s equation along the interface of the gravity current near the head, we end up with a front condition that is in better agreement with experiment than previous theoretical models. This condition is a function of the density ratio between current and ambient fluids, which was different from previous theoretical models, while it has been widely used in experimental studies. The present front condition suggests that the form drag may account for a significant part of the resistance force applied on the current head.

Stability of the inviscid plane couette flow in bubbly fluids

Abstract: The aim of this article is to study the stability of shear flows in bubbly fluids. A mathematical model of bubbly fluids is presented. The stability of shear flows is studied by two methods: by using a spectral approach and by solving the initial-value problem. It is proved that the linear velocity profile is stable in the long wave approximation.

An infinitude of constant-energy flows of a gas

Abstract: By converting the usual equations of motion for ideal inviscid fluids to Lagrangian form it is possible to exhibit infinitely many constant-energy flows of a compressible gas.