The dynamics of fluid–fluid interfaces are important in diverse problems that span many disciplines in science and engineering. A series of snapshots is used to illustrate the breadth of applications that can occur in viscous low-Reynolds-number flows and I highlight theoretical and modelling ideas that are broadly useful for these, as well as other, problems. By way of illustration of unifying quantitative ideas we discuss briefly (i) the use of the Reciprocal Theorem in low-Reynolds-number flows, (ii) the use of the lubrication approximation for characterizing thin-film coating flows sometimes referred to as Landau–Levich–Derjaguin–Bretherton problems and (iii) nearly two-dimensional viscously dominated flows.

In the literature it has so far been common practice to consider solitary waves and N-waves (composed of solitary waves) as the appropriate model of tsunamis approaching the shoreline. Unfortunately, this approach is based on a tie between the nonlinearity and the horizontal length scale (or duration) of the wave, which is not realistic for geophysical tsunamis. To resolve this problem, we first derive analytical solutions to the nonlinear shallow-water (NSW) equations for the runup/rundown of single waves, where the duration and the wave height can be specified separately. The formulation is then extended to cover leading depression N-waves composed of a superposition of positive and negative single waves. As a result the temporal variations of the runup elevation, the associated velocity and breaking criteria are specified in terms of polylogarithmic functions. Finally, we consider incoming transient wavetrains generated by monopole and dipole disturbances in the deep ocean. The evolution of these wavetrains, while travelling a considerable distance over a constant depth, is influenced by weak dispersion and is governed by the linear Korteweg–De Vries (KdV) equation. This process is described by a convolution integral involving the Airy function. The runup on the plane sloping beach is then determined by another convolution integral involving the incoming time series at the foot of the slope. A good agreement with numerical model results is demonstrated.

In this work we explore if it is possible to reproduce molecular-scale slip behaviour by using continuum equations. To that end it is noted that molecular-scale slip is affected by three factors: (i) near the wall, the fluid experiences a potential because of the wall; (ii) the fluid density responds to that potential, and hence, fluid compressibility is relevant; and (iii) the fluid can lose momentum to the wall. To incorporate these features we simulate shear flow of a compressible fluid between two walls in the presence of a potential. Compressibility effect is found to be important only in the near-wall region. The slip length is calculated from the mean velocity profile. The slip-length-versus-shear-rate trend is similar to that in molecular dynamic calculations. First, there is a constant value of the slip length at low shear rates. Then, the slip length increases beyond a critical shear rate. Lastly, the slip length reaches another constant value if the wall momentum loss parameter is non-zero. The scaling for the critical shear rate emerges from our results. The value of the slip length increases if the wall potential is less corrugated and if the momentum loss to the wall is low. An understanding of the overall force balance during various slip modes emerges from the governing equations.

This paper theoretically investigates the influence of the upstream boundary conditions on the bifurcation structure leading to vortex breakdown. The axisymmetric flow of an inviscid fluid in a pipe of constant cross-section and finite axial length is considered. Solutions bifurcating from the columnar solution at criticality are analysed via a weakly nonlinear expansion and computed in the fully nonlinear regime using numerical continuation, until a centreline recirculation is found at the pipe outlet. Bifurcation diagrams are determined for a parametric family of inflows describing a wide range of axial and azimuthal profiles, the third inlet condition being chosen either as a fixed azimuthal vorticity or as a vanishing radial velocity. Including the traditional picture given by Wang & Rusak (J. Fluid Mech., vol. 340, 1997a, p. 177), six different diagrams are found to be possible. In particular, a scenario of smooth transition to breakdown may exist as the swirl is increased, with no loss of stability and no hysteresis, breakdown appearing for swirl levels larger than the critical swirl in a pipe. This transition involves a new type of flow akin to a pre-breakdown flow. Our results, furthermore, suggest that rigidly rotating Poiseuille flow could correspond to the limit for which breakdown is impossible because it is predicted at infinitely large swirl numbers. We finally find that flows with a large rotational core are particularly sensitive to an accurate modelling of the upstream boundary conditions, weakly confined vortices being much more robust.

An analytical model is developed to explain the equilibration mechanism of the salt finger instability in unbounded temperature and salinity gradients. The theory is based on the weakly nonlinear asymptotic expansion about the point of marginal instability. The proposed solutions attribute equilibration of salt fingers to a combination of two processes: (i) the triad interaction and (ii) spontaneous development of the mean vertical shear. The non-resonant triad interactions control the equilibration of linear growth for moderate and large values of Prandtl number (Pr) and for slightly unstable parameters. For small Pr and/or rigorous instabilities, the mean shear effects become essential. It is shown that, individually, neither the mean field nor the triad interaction models can accurately describe the equilibrium patterns of salt fingers in all regions of the parameter space. Therefore, we propose a new hybrid model, which represents both stabilizing effects in a single framework. The resulting solutions agree with the fully nonlinear numerical simulations over a wide range of governing parameters.

We consider the steady axisymmetric motion of an electrically conducting fluid contained within a spherical shell and permeated by a centred axial dipole magnetic field, which is strong as measured by the Hartmann number M. Slow axisymmetric motion is driven by rotating the inner boundary relative to the stationary outer boundary. For M ≫ 1, viscous effects are only important in Hartmann boundary layers adjacent to the inner and outer boundaries and a free shear-layer on the magnetic field line that is tangent to the outer boundary on the equatorial plane of symmetry. We measure the ability to leak electric current into the solid boundaries by the size of their relative conductance ɛ. Since the Hartmann layers are sustained by the electric current flow along them, the current inflow from the fluid mainstream needed to feed them increases in concert with the relative conductance, because of the increasing fraction ℒ of the current inflow leaked directly into the solids. Therefore the nature of the flow is sensitive to the relative sizes of ɛ−1 and M.

The current work extends an earlier study of the case of a conducting inner boundary and an insulating outer boundary with conductance ɛo = 0 (Dormy, Jault & Soward, J. Fluid Mech., vol. 452, 2002, pp. 263–291) to other values of the outer boundary conductance. Firstly, analytic results are presented for the case of perfectly conducting inner and outer boundaries, which predict super-rotation rates Ωmax of order M1/2 in the free shear-layer. Successful comparisons are made with numerical results for both perfectly and finitely conducting boundaries. Secondly, in the case of a finitely conducting outer boundary our analytic results show that Ωmax is O(M1/2) for ɛo−1 ≪ 1 ≪ M3/4, O(ɛo2/3M1/2) for 1 ≪ ɛo−1 ≪ M3/4 and O(1) for 1 ≪ M3/4 ≪ ɛo−1. On increasing ɛo−1 from zero, substantial electric current leakage into the outer boundary, ℒo ≈ 1, occurs for ɛo−1 ≪ M3/4 with the shear-layer possessing the character appropriate to a perfectly conducting outer boundary. When ɛo−1 = O(M3/4) the current leakage is blocked near the equator, and the nature of the shear-layer changes. So, when M3/4 ≪ ɛo−1, the shear-layer has the character appropriate to an insulating outer boundary. More precisely, over the range M3/4 ≪ ɛo−1 ≪ M the blockage spreads outwards, reaching the pole when ɛo−1 = O(M). For M ≪ ɛo−1 current flow into the outer boundary is completely blocked, ℒo ≪ 1.

The thin-Debye-layer model is utilized to analyse the electro-hydrodynamic flow about a colloidal particle which is exposed to a direct ionic current, emitted by a proximate reactive electrode. This flow is driven by electro-osmotic slip on the particle, as well as a comparable slip on the electrode itself. The small particle–electrode separation allows for the use of singular perturbation. Thus, the electro-neutral bulk-fluid domain is decomposed into an ‘inner’ gap region, where the electric field and shear rate are large, and an ‘outer’ region, consisting of the remaining bulk domain, where they are moderate. Matched asymptotic expansions in both regions provide the requisite flow field. The intensive shear rate in the narrow gap region is associated with a lubrication-type pressure build-up, which is responsible for the leading-order hydrodynamic force on the particle. This force acts to repel the particle away from the electrode, thereby supporting it against gravity. Its magnitude is inversely proportional to the gap width. At large distances from the particle the fluid velocity decays with the third power of distance, while near the electrode it decays with the fourth power. The inward pointing flow near the electrode tend to entrain neighbouring particles, thereby resulting in two-dimensional particle clusters. For equal values of particle and anode zeta potentials, this process is dominated by the particle-slip contribution.

A large eddy simulation (LES) a posteriori study is conducted for a temporal mixing layer which initially contains different species in the lower and upper streams and in which the initial pressure is larger than the critical pressure of either species. A vorticity perturbation, initially imposed, promotes roll-up and a double pairing of four initial spanwise vortices to reach a transitional state. The LES equations consist of the differential conservation equations coupled with a real-gas equation of state, and the equations utilize transport properties depending on the thermodynamic variables. Unlike all LES models to date, the differential equations contain, additional to the subgrid-scale (SGS) fluxes, a new SGS term denoted a ‘pressure correction’ (p correction) in the momentum equation. This additional term results from filtering the Navier–Stokes equations and represents the gradient of the difference between the filtered p and p computed from the filtered flow field. A previous a priori analysis, using a direct numerical simulation (DNS) database for the same configuration, found this term to be of leading order in the momentum equation, a fact traced to the existence of regions of high density-gradient magnitude that populated the entire flow; in that study, the appropriateness of several SGS-flux models was assessed, and a model for the p-correction term was proposed.

In the present study, the constant-coefficient SGS-flux models of the a priori investigation are tested a posteriori in LES devoid of, or including, the SGS p-correction term. A new p-correction model, different from that of the a priori study, is used, and the results of the two p-correction models are compared. The results reveal that the former is less computationally intensive and more accurate than the latter in reproducing global and structural features of the flow. The constant-coefficient SGS-flux models encompass the Smagorinsky (SMC) model, in conjunction with the Yoshizawa (YO) model for the trace, the gradient (GRC) model and the scale similarity (SSC) models, all exercised with the a priori study constant-coefficient values calibrated at the transitional state. Further, dynamic SGS-flux model LESs are performed with the p correction included in all cases. The dynamic models are the Smagorinsky (SMD) model, in conjunction with the YO model, the gradient (GRD) model and ‘mixed’ models using SMD in combination with GRC or SSC utilized with their theoretical coefficient values. The LES comparison is performed with the filtered-and-coarsened DNS (FC-DNS) which represents an ideal LES solution. The constant-coefficient models including the p correction (SMCP, GRCP and SSCP) are substantially superior to those devoid of it; the SSCP model produces the best agreement with the FC-DNS template. For duplicating the local flow structure, the predictive superiority of the dynamic mixed models is demonstrated over the SMD model; however, even better predictions in capturing vortical features are obtained with the GRD model. The GRD predictions improve when LES is initiated at a time past the initial range in which the p-correction term rivals in magnitude the leading-order term in the momentum equation. Finally, the ability of the LES to predict the FC-DNS irreversible entropy production is assessed. It is shown that the SSCP model is the best at recovering the domain-averaged irreversible entropy production. The sensitivity of the predictions to the initial conditions and grid size is also investigated.

In this work, we analyse the linear stability of a frozen Lamb–Oseen vortex in a fluid linearly stratified along the vortex axis. The temporal stability properties of three-dimensional normal modes are obtained under the Boussinesq approximation with a Chebychev collocation spectral code for large ranges of Froude numbers and Reynolds numbers (the Schmidt number being fixed to 700). A specific integration technique in the complex plane is used in order to apply the condition of radiation at infinity. For large Reynolds numbers and small Froude numbers, we show that the vortex is unstable with respect to all non-axisymmetrical waves. The most unstable mode is however always a helical radiative mode (m = 1) which resembles either a displacement mode or a ring mode. The displacement mode is found to be unstable for all Reynolds numbers and for moderate Froude numbers (F ~ 1). The radiative ring mode is by contrast unstable only for large Reynolds numbers above 104 and is the most unstable mode for large Froude numbers (F > 2). The destabilization of this mode for large Froude numbers is shown to be associated with a resonance mechanism which is analysed in detail. Analyses of the scaling and of the spatial structure of the different unstable modes are also provided.

Diffuse-interface models may be used to compute moving contact lines because the Cahn–Hilliard diffusion regularizes the singularity at the contact line. This paper investigates the basic questions underlying this approach. Through scaling arguments and numerical computations, we demonstrate that the Cahn–Hilliard model approaches a sharp-interface limit when the interfacial thickness is reduced below a threshold while other parameters are fixed. In this limit, the contact line has a diffusion length that is related to the slip length in sharp-interface models. Based on the numerical results, we propose a criterion for attaining the sharp-interface limit in computing moving contact lines.

Gravitational available potential energy is a central concept in an energy analysis of flows in which buoyancy effects are dynamically important. These include, but are not limited to, most geophysical flows with persistently stable density stratification. The volume-integrated available potential energy ap is defined as the difference between the gravitational potential energy of the system and the potential energy of a reference state with the lowest potential energy that can be reached by adiabatic material rearrangement; ap determines how much energy is available for conservative dynamical exchange with kinetic energy k. In this paper we introduce new techniques for computing the local available potential energy density Eap in numerical simulations that allow for a more accurate and complete analysis of the available potential energy and its dynamical balances as part of the complete energy cycle of a flow. In particular, the definition of Eap and an associated gravitation disturbance field permit us to make a spectral decomposition of its dynamical balance and examine the cross-scale energy flux. Several examples illustrate the spatial structure of Eap and its evolutionary influences. The greatest attention is given to an analysis of a turbulent-equilibrium simulation Eady-like vertical-shear flow with rotation and stable stratification. In this regime Eap exhibits a vigorous forward energy cascade from the mesoscale through the submesoscale range – first in a scale range dominated by frontogenesis and positive buoyancy-flux conversion from ap to k and then, after strong frontal instability and frontogenetic arrest, in a coupled kinetic-potential energy inertial-cascade range with negative buoyancy-flux conversion – en route to fine-scale dissipation of both energy components.

The generation and energy downscaling of wind-forced long internal waves in strongly stratified small-to-medium sized narrow lakes are studied. A two-layer nonlinear model with forcing and damping is employed. Even though the wave field is fundamentally bidirectional in nature, a domain folding technique is employed to simulate the leading-order internal wave field in terms of a weakly nonlinear weakly dispersive model equation of Korteweg–deVries type. Parametric effects of wind-forcing and environmental conditions, including variable topography and variable basin width, are examined. Energy downscaling from basin-scale waves to shorter scales are quantified by means of a time evolution of the wave energy spectra. It is demonstrated that an internal wave resonance is possible when repetitive wind-forcing events arise with a frequency near the linear seiche frequency. An attempt is made to apply the model to describe the shoaling of long waves on sloping endwall boundaries. Modelling of the energy loss and energy reflection during a shoaling event is calibrated against laboratory experiments.

Numerical simulations are used to investigate the flow associated with a bluegill sunfish (Lepomis macrochirus) pectoral fin during steady forward motion. The simulations are intended to match the experiments of Lauder et al. (Bioinsp. Biomim., vol. 1, 2006, p. S25), and the results obtained from the simulations complement the experimental analysis. The focus of the current paper is on the quantitative characterization of the propulsive performance of the pectoral fin, which undergoes significant deformation during its stroke. This includes a detailed analysis of the thrust production mechanisms as well as their connection to the vortex dynamics and other flow features. The simulations indicate that the fish fin produces high propulsive performance by employing a complex fin gait driven by active and passive fin deformation. By connecting the vortex dynamics and fin kinematics with the surface distribution of the force on the fin, it is found that during abduction, the fin moves such that the tip of the fin undergoes a complex, three-dimensional flapping motion that produces a strong and long-lasting, attached tip vortex. This tip vortex is associated with most of the thrust production during the abduction phase of the stroke. During the adduction phase, the fin motion is similar to a ‘paddling’ stroke. Comparisons are made with rigid flapping foils to provide insights into the remarkable performance of the fish fin and to interpret the force production from the viewpoint of functional morphology.

We present a theoretical stability analysis of the flow after the sudden release of a fixed mass of fluid on an inclined plane formally restricted to relatively long time scales, for which the kinematic regime is valid. Shallow-water equations for steep slopes with bed stress are employed to study the threshold for the onset of roll waves. An asymptotic solution for long-wave perturbations of small amplitude is found on background flows with a Froude number value of 2. Small disturbances are stable under this condition, with a linear decay rate independent of the wavelength and with a wavelength that increases linearly with time. For larger values of the Froude number it is shown that the basic flow moves at a different scale than the perturbations, and hence the wavelength of the unstable modes is characterized as a function of the plane-parallel Froude number Frp and a measure of the local slope of the free-surface height φ by means of a multiple-scale analysis in space and time. The linear stability results obtained in the presence of small non-uniformities in the flow, φ > 0, introduce substantial differences with respect to the plane-parallel flow with φ = 0. In particular, we find that instabilities do not occur at Froude numbers Frcr much larger than the critical value 2 of the parallel case for some wavelength ranges. These results differ from that previously reported by Lighthill & Whitham (Proc. R. Soc. A, vol. 229, 1955, pp. 281–345), because of the fundamental role that the non-parallel, time-dependent characteristics of the kinematic-wave play in the behaviour of small disturbances, which was neglected in their stability analyses. The present work concludes with supporting numerical simulations of the evolution of small disturbances, within the framework of the frictional shallow-water equations, that are superimposed on a base state which is essentially a kinematic wave, complementing the asymptotic theory relevant near the onset. The numerical simulations corroborate the cutoff in wavelength for the spectrum that stabilizes the tail of the dam-break flood.

We study nonlinear, two-dimensional convection in a mushy layer during solidification of a binary mixture. We consider a particular limit in which the onset of oscillatory convection just precedes the onset of steady overturning convection, at a prescribed aspect ratio of convection patterns. This asymptotic limit allows us to determine nonlinear solutions analytically. The results provide a complete description of the stability of and transitions between steady and oscillatory convection as functions of the Rayleigh number and the compositional ratio. Of particular focus are the effects of the basic-state asymmetries and non-uniformity in the permeability of the mushy layer, which give rise to abrupt (hysteretic) transitions in the system. We find that the transition between travelling and standing waves, as well as that between standing waves and steady convection, can be hysteretic. The relevance of our theoretical predictions to recent experiments on directionally solidifying mushy layers is also discussed.

A series of direct numerical simulations, both in two- and three-dimensions, of the flow past a circular cylinder for Reynolds numbers Re ≤ 600 has been conducted. From these simulations, the time-mean (and, for the three-dimensional simulations, the spanwise spatial-mean) flow has been calculated. A global linear stability analysis has been conducted on these mean flows, showing that the mean cylinder wake for Re ≤ 600 is marginally stable and the eigenfrequency of the leading global mode closely predicts the eventual saturated vortex shedding frequency. A local stability analysis has also been conducted. For this, a series of streamwise velocity profiles has been extracted from the mean wake and the stability of these profiles has been analysed using the Rayleigh stability equation. The real and imaginary instability frequencies gained from these profiles have then been used to find the global frequency selected by the flow using a saddle-point criterion. The results confirm the success of the saddle-point criterion when the mean flow is quasi-parallel in the vicinity of the saddle point; however, the limitations of the method when the mean flow exhibits higher curvature are also elucidated.

We present an estimator-based control design procedure for flow control, using reduced-order models of the governing equations linearized about a possibly unstable steady state. The reduced-order models are obtained using an approximate balanced truncation method that retains the most controllable and observable modes of the system. The original method is valid only for stable linear systems, and in this paper, we present an extension to unstable linear systems. The dynamics on the unstable subspace are represented by projecting the original equations onto the global unstable eigenmodes, assumed to be small in number. A snapshot-based algorithm is developed, using approximate balanced truncation, for obtaining a reduced-order model of the dynamics on the stable subspace.

The proposed algorithm is used to study feedback control of two-dimensional flow over a flat plate at a low Reynolds number and at large angles of attack, where the natural flow is vortex shedding, though there also exists an unstable steady state. For control design, we derive reduced-order models valid in the neighbourhood of this unstable steady state. The actuation is modelled as a localized body force near the trailing edge of the flat plate, and the sensors are two velocity measurements in the near wake of the plate. A reduced-order Kalman filter is developed based on these models and is shown to accurately reconstruct the flow field from the sensor measurements, and the resulting estimator-based control is shown to stabilize the unstable steady state. For small perturbations of the steady state, the model accurately predicts the response of the full simulation. Furthermore, the resulting controller is even able to suppress the stable periodic vortex shedding, where the nonlinear effects are strong, thus implying a large domain of attraction of the stabilized steady state.

Low-frequency waves that develop in a shallow layer of fluid, contained in a channel with linearly slopping bottom and rotating with uniform angular speed are investigated theoretically and experimentally. Exact numerical solutions of the eigenvalue problem, obtained from the linearized shallow water equations on the f-plane, show that the waves are trapped near the channel's shallow wall and propagate along it with the shallow side on their right in the Northern hemisphere. The phase speed of the waves is slower compared with that of the harmonic theory in which bottom slope is treated inconsistently. A first-order approximation of the cross-channel dependence of the coefficient in the eigenvalue equation yields an approximation of the cross-channel velocity eigenfunction as an Airy function, which, for sufficiently wide channels, yields an explicit expression for the wave's dispersion relation. The analytic solutions of the eigenvalue problem agree with the numerical solutions in both the wave trapping and the reduced phase speed. For narrow channels, our theory yields an estimate of the channel width below which the harmonic theory provides a more accurate approximation. Laboratory experiments were conducted on a 13 m diameter turntable at LEGI-Coriolis (France) into which a linearly sloping bottom of 10 % incline was installed. A wavemaker generated waves of known frequency at one end of the turntable and the wavenumbers of these waves were measured at the opposite end using a particle imaging velocimetry technique. The experimental results regarding the phase speed and the radial structure of the amplitude are in very good agreement with our theoretical non-harmonic predictions, which support the present modification of the harmonic theory in wide channels.

The relative dispersion of pairs of inertial point particles in incompressible, homogeneous and isotropic three-dimensional turbulence is studied by means of direct numerical simulations at two values of the Taylor-scale Reynolds number Reλ ~ 200 and Reλ ~ 400, corresponding to resolutions of 5123 and 20483 grid points, respectively. The evolution of both heavy and light particle pairs is analysed by varying the particle Stokes number and the fluid-to-particle density ratio. For particles much heavier than the fluid, the range of available Stokes numbers is St ∈ [0.1 : 70], while for light particles the Stokes numbers span the range St ∈ [0.1 : 3] and the density ratio is varied up to the limit of vanishing particle density. For heavy particles, it is found that turbulent dispersion is schematically governed by two temporal regimes. The first is dominated by the presence, at large Stokes numbers, of small-scale caustics in the particle velocity statistics, and it lasts until heavy particle velocities have relaxed towards the underlying flow velocities. At such large scales, a second regime starts where heavy particles separate as tracers' particles would do. As a consequence, at increasing inertia, a larger transient stage is observed, and the Richardson diffusion of simple tracers is recovered only at large times and large scales. These features also arise from a statistical closure of the equation of motion for heavy particle separation that is proposed and is supported by the numerical results. In the case of light particles with high density ratio, strong small-scale clustering leads to a considerable fraction of pairs that do not separate at all, although the mean separation increases with time. This effect strongly alters the shape of the probability density function of light particle separations.

Transition to turbulence in a pipe is characterized by the increase of the characteristic lifetimes of localized turbulent spots (‘puffs’) with increasing Reynolds number (Re). Previous experiments are based on visualization or indirect measurements of the lifetime probability. Here we report quantitative direct measurements of the lifetimes based on accurate pressure measurements combined with laser Doppler anemometry (LDA). The characteristic lifetime is determined directly from the lifetime probability. It is shown that the characteristic lifetime does not diverge at finite Re, and follows an exponential scaling for the observed range 1725 ≤ Re ≤ 1955. Over this small Re range the lifetime increases over four orders of magnitude. The results show that the puff velocity is not constant, and the rapid disintegration of puffs occurs within 20–70 pipe diameters.