Turbulence is widely considered one of the most important and most difficult unsolved problems in classical physics. It is also the area of fluid mechanics where the greatest effort is exerted, the most papers published and, some would argue, the least progress made. Although direct numerical simulation is becoming an increasingly valuable tool, there remains a need for high-quality experiments to underpin our theoretical and numerical progress. Such statements apply equally to the ‘classical’ problem of homogeneous isotropic turbulence and to turbulence in its many other guises. Of particular interest is turbulence in a rotating system, where it is well known that the influence of rotation leads to the development of anisotropy and the elongation of scales parallel to the rotation axis. Moisy et al. (J. Fluid Mech., 2010, this issue, vol. 666, pp. 5–35) present new experiments in the free decay of grid-generated turbulence in a rotating system. They investigate the emergence of anisotropy from essentially isotropic initial conditions. While it is well known that rotation suppresses velocity gradients parallel to the rotation axis, Moisy et al. (2010) uncover some startling and previously overlooked implications.

The effect of a background rotation on the decay of grid-generated turbulence is investigated from experiments using the large-scale ‘Coriolis’ rotating platform. A first transition occurs at 0.4 tank rotation (instantaneous Rossby number Ro ≃ 0.25), characterized by a t−6/5 → t−3/5 transition of the energy-decay law. After this transition, anisotropy develops in the form of vertical layers, where the initial vertical velocity fluctuations remain trapped. The vertical vorticity field develops a cyclone–anticyclone asymmetry, reproducing the growth law of the vorticity skewness, Sω(t) ≃ (Ωt)0.7, reported by Morize, Moisy & Rabaud (Phys. Fluids, vol. 17 (9), 2005, 095105). A second transition is observed at larger time, characterized by a return to vorticity symmetry. In this regime, the layers of nearly constant vertical velocity become thinner as they are advected and stretched by the large-scale horizontal flow, and eventually become unstable. The present results indicate that the shear instability of the vertical layers contributes significantly to the re-symmetrization of the vertical vorticity at large time, by re-injecting vorticity fluctuations of random sign at small scales. These results emphasize the importance of the nature of the initial conditions in the decay of rotating turbulence.

Motivated by the need to gain fundamental insights into the mechanisms of bed-load sediment transport in turbulent junction flows, we carry out a computational study of Lagrangian dynamics of inertial particles initially placed on the bed upstream of a surface-mounted circular cylinder in a rectangular open channel (Dargahi, J. Hydraul. Engng, vol. 116, 1990, pp. 1197–1214). The flow field at Re = 39000 is simulated using the detached eddy simulation (DES) approach (Spalart et al., In Advances in DNS/LES, ed. C. Liu & Z. Liu, 1997, Greyden), which has already been shown to accurately resolve most of the turbulent stresses produced by the low-frequency, bimodal fluctuations of the turbulent horseshoe vortex (Paik et al., J. Hydraul. Engng, vol. 131, 1990, pp. 441–456; Escauriaza & Sotiropoulos, Flow Turbul. Combust., 2010, in press). The trajectory and momentum equations for the sediment particles are integrated numerically simultaneously with the flow governing equations assuming one-way coupling and neglecting particle-to-particle interactions (dilute flow) but taking into account bed–particle interactions and the effects of the instantaneous hydrodynamic forces induced by the resolved fluctuations of the coherent vortical structures. The computed results show that, in accordance with the simulated clear-water scour condition (i.e. the magnitude of the particle stresses is near the threshold of motion), the transport of sediment grains is highly intermittent and exhibits essentially all the characteristics of bed-load sediment transport observed in laboratory and field experiments. Groups of sediment grains are dislodged from the bed simultaneously in seemingly random bursting events and begin to move, saltating or sliding along the bed. Furthermore, particles that are not entrained into the bed-load layer are found to form streaks aligned with near-wall vortices around the cylinder. The global transport of particles is studied by performing a statistical analysis of the bed-load flux to reveal scale-invariance of the process and multifractality of particle transport as the overall effect of the coherent structures of the flow. A major finding of this work is that a relatively simple Lagrangian model coupled with a coherent-structure resolving simulation of the turbulent flow is able to reproduce the sediment dynamics observed in multiple experiments performed under similar conditions, and provide fundamental information on the initiation of motion and the multifractal nature of bed-load transport processes. The results also motivate the development of new Eulerian bed-load transport models that consider unsteady conditions and incorporate the intermittency of the unresolved scales of sediment motion.

The Honji instability is studied using direct numerical simulations of sinusoidal oscillatory flow around a circular cylinder. The three-dimensional Navier–Stokes equations are solved by a finite element method at a relatively small value of the Keulegan–Carpenter number KC. The generation and subsequent development of Honji vortices are discussed over a range of frequency parameters by means of flow visualization. It is found that the spacing between Honji vortices is only weakly dependent on the frequency of oscillation, but is strongly correlated to KC because it is the terms within the governing equation containing KC that dominate the three-dimensional features of the flow. An empirical relationship between KC and the spacing between neighbouring vortices is proposed. The three-dimensional steady streaming structure within the vortices is identified and it is found that at high frequencies the steady streaming is two-dimensional although the instantaneous flow structure is itself fully three-dimensional.

In this paper, the instability of a fluid inside a precessing cylinder is addressed theoretically and experimentally. The precessional motion forces Kelvin modes in the cylinder, which can become resonant for given precessional frequencies and cylinder aspect ratios. When the Reynolds number is large enough, these forced resonant Kelvin modes eventually become unstable. A linear stability analysis based on a triadic resonance between a forced Kelvin mode and two additional free Kelvin modes is carried out. This analysis allows us to predict the spatial structure of the instability and its threshold. These predictions are compared to the vorticity field measured by particle image velocimetry with an excellent agreement. When the Reynolds number is further increased, nonlinear effects appear. A weakly nonlinear theory is developed semi-empirically by introducing a geostrophic mode, which is triggered by the nonlinear interaction of a free Kelvin mode with itself in the presence of viscosity. Amplitude equations are obtained coupling the forced Kelvin mode, the two free Kelvin modes and the geostrophic mode. They show that the instability saturates to a fixed point just above threshold. Increasing the Reynolds number leads to a transition from a steady saturated regime to an intermittent flow in good agreement with experiments. Surprisingly, this weakly nonlinear model still gives a correct estimate of the mean flow inside the cylinder even far from the threshold when the flow is turbulent.

We study the capillary rise of wetting liquids in the corners of different geometries and show that the meniscus rises without limit following the universal law: h(t)/a ≈ (γt/ηa)1/3, where γ and η stand for the surface tension and viscosity of the liquid while is the capillary length, based on the liquid density ρ and gravity g. This law is universal in the sense that it does not depend on the geometry of the corner.

The dynamics of a plane interface separating two sheared, density and viscosity matched fluids in the vertical gap between parallel plate electrodes are studied computationally. A Couette profile is imposed onto the fluids by moving the rigid plates at equal speeds in opposite directions. In addition, a vertical electric field is applied to the shear flow by impressing a constant voltage difference on the electrodes. The stability of the initially flat interface is a very subtle balance between surface tension, inertia, viscosity and electric field effects. Under unstable conditions, the potential difference in the fluid results in an electrostatic pressure that amplifies disturbance waves on the two-fluid interface at characteristic wave lengths. Various mechanisms determining the growth rate of the most unstable mode are addressed in a systematic parameter study. The applied methodology involves a combination of numerical simulation and analytical work. Linear stability theory is employed to identify unstable parametric conditions of the perturbed Couette flow. Particular attention is given to the effect of the applied electric field on the instability of the perturbed two-fluid interface. The normal mode analyses are followed up by numerical simulations. The applied method relies on solving the governing equations for the fluid mechanics and the electrostatics in a one-fluid approximation by using a finite-volume technique combined with explicit tracking of the evolving interface. The numerical results confirm those of linear theory and, furthermore, reveal a rich array of dynamical behaviour. The elementary fluid instabilities are finger-like structures of interpenetrating fluids. For weakly unstable situations a single fingering instability emerges on the interface. Increasing the growth rates causes the finger to form a drop-like tip region connected by a long thinning fluids neck. Even more striking fluid motion occurs at higher values of the electric field parameter for which multiple fluid branches develop on the interface. For a pair of perfect dielectrics the vertical electric field was found to enhance interfacial motion irrespective of the permittivity ratio, while in leaky dielectrics the electric field can either stabilize or destabilize the interface, depending on the conductivity and permittivity ratio between the fluids.

This paper explores the phenomena associated with the emergence of gas bubbles from a submerged granular bed. While there are many natural and industrial applications, we focus on the particular circumstances and consequences associated with the emergence of methane bubbles from the beds of lakes and reservoirs since there are significant implications for the dynamics of lakes and reservoirs and for global warming. This paper describes an experimental study of the processes of bubble emergence from a granular bed. Two distinct emergence modes are identified, mode 1 being simply the percolation of small bubbles through the interstices of the bed, while mode 2 involves the cumulative growth of a larger bubble until its buoyancy overcomes the surface tension effects. We demonstrate the conditions dividing the two modes (primarily the grain size) and show that this accords with simple analytical evaluations. These observations are consistent with previous studies of the dynamics of bubbles within porous beds. The two emergence modes also induce quite different particle fluidization levels. The latter are measured and correlated with a diffusion model similar to that originally employed in river sedimentation models by Vanoni and others. Both the particle diffusivity and the particle flux at the surface of the granular bed are measured and compared with a simple analytical model. These mixing processes can be consider applicable not only to the grains themselves, but also to the nutrients and/or contaminants within the bed. In this respect they are shown to be much more powerful than other mixing processes (such as the turbulence in the benthic boundary layer) and could, therefore, play a dominant role in the dynamics of lakes and reservoirs.

A weakly nonlinear theory, in terms of the well-known Landau equation, has been developed to describe the nonlinear saturation of the shear-banding instability in a rapid granular plane Couette flow using the amplitude expansion method. The nonlinear modes are found to follow certain symmetries of the base flow and the fundamental mode, which helped to identify analytical solutions for the base-flow distortion and the second harmonic, leading to an exact calculation of the first Landau coefficient. The present analytical solutions are used to validate a spectral-based numerical method for the nonlinear stability calculation. The regimes of supercritical and subcritical bifurcations for the shear-banding instability have been identified, leading to the prediction that the lower branch of the neutral stability contour in the (H, φ0)-plane, where H is the scaled Couette gap (the ratio between the Couette gap and the particle diameter) and φ0 is the mean density or the volume fraction of particles, is subcritically unstable. The predicted finite-amplitude solutions represent shear localization and density segregation along the gradient direction. Our analysis suggests that there is a sequence of transitions among three types of pitchfork bifurcations with increasing mean density: from (i) the bifurcation from infinity in the Boltzmann limit to (ii) subcritical bifurcation at moderate densities to (iii) supercritical bifurcation at larger densities to (iv) subcritical bifurcation in the dense limit and finally again to (v) supercritical bifurcation near the close packing density. It has been shown that the appearance of subcritical bifurcation in the dense limit depends on the choice of the contact radial distribution function and the constitutive relations. The scalings of the first Landau coefficient, the equilibrium amplitude and the phase diagram, in terms of mode number and inelasticity, have been demonstrated. The granular plane Couette flow serves as a paradigm that supports all three possible types of pitchfork bifurcations, with the mean density (φ0) being the single control parameter that dictates the nature of the bifurcation. The predicted bifurcation scenario for the shear-band formation is in qualitative agreement with particle dynamics simulations and the experiment in the rapid shear regime of the granular plane Couette flow.

Experiments in small aspect-ratio Taylor–Couette flows have reported the presence of a band in parameter space where rotating waves become steady non-axisymmetric solutions (a pinning effect) via infinite-period bifurcations. Previous numerical simulations were unable to reproduce these observations. Recent additional experiments suggest that the pinning effect is not intrinsic to the dynamics of the problem, but rather is an extrinsic response induced by the presence of imperfections. Here we present numerical simulations that include a small tilt of one of the endwalls, simulating the effects of imperfections that break the SO(2) axisymmetry of the problem, and indeed are able to reproduce the experimentally observed pinning of the rotating waves. Dynamical systems considerations suggest that any imperfection breaking the SO(2) axisymmetry of the problem must result in the formation of a pinning region of finite width. We have also found that the particulars of the pinning process, in particular the width of the pinning region, are extremely sensitive to the type of imperfection in the system. Almost identical flows respond in completely different ways to the same imperfection, depending on subtle differences in the weak secondary characteristics of the flow. The numerical simulations of the Navier–Stokes equations for the problem with an imposed tilt of an endwall together with normal-form analysis of a Hopf bifurcation subjected to imposed symmetry-breaking help shed some light on previous experiments that reported a variety of different dynamical behaviour for which a clear explanation was lacking.

We report on the direct experimental observation of laser-induced flows in isotropic liquids that scatter light. We use a droplet microemulsion in the two-phase regime, which behaves like a binary mixture. Close to its critical consolute line, the microemulsion undergoes large refractive index fluctuations that scatter light. The radiation pressure of a laser beam is focused onto the soft interface between the two phases of the microemulsion and induces a cylindrical liquid jet that continuously emits droplets. We demonstrate that this dripping phenomenon takes place as a consequence of a steady flow induced by the transfer of linear momentum from the optical field to the liquid due to light scattering. We first show that the cylindrical jet guides light as a step-index liquid optical fiber whose core diameter is self-adapted to the light itself. Then, by modelling the light-induced flow as a low-Reynolds-number, parallel flow, we predict the dependence of the dripping flow rate on the thermophysical properties of the microemulsion and the laser beam power. Satisfying agreement is found between the model and experiments.

A joint theoretical and experimental study is performed on the generation of internal gravity waves by an oscillating sphere, as a paradigm of the generation of internal tides by barotropic tidal flow over three-dimensional supercritical topography. The theory is linear and three-dimensional, applies both near and far from the sphere, and takes into account viscosity and the unsteadiness arising from the interference with transients generated at the start-up. The waves propagate in conical beams, evolving with distance from a bimodal to unimodal wave profile. In the near field, the profile is asymmetric with its major peak towards the axis of the cones. The experiments involve horizontal oscillations and develop a cross-correlation technique for the measurement of the deformation of fluorescent dye planes to sub-pixel accuracy. At an oscillation amplitude of one fifth of the radius of the sphere, the waves are linear and the agreement between experiment and theory is excellent. As the amplitude increases to half the radius, nonlinear effects cause the wave amplitude to saturate at a value that is 20% lower than its linear estimate. Application of the theory to the conversion rate of barotropic tidal energy into internal tides confirms the expected scaling for flat topography, and shows its transformation for hemispherical topography. In the ocean, viscous and unsteady effects have an essentially local role, in keeping the wave amplitude finite at the edges of the beams, and otherwise dissipate energy on such large distances that they hardly induce any decay.

Series solutions are found which describe the evolution to absolute and convective instability in an inviscid liquid sheet flowing in a quiescent ambient gas and subject to a localized perturbation. These solutions are used to validate asymptotic stability predictions for sinuous and varicose disturbances. We show how recent disagreements in growth predictions stem from assumptions made when arriving at the Fourier integral response. Certain initial conditions eliminate or reduce the order of singularities in the Fourier integral. If a Gaussian perturbation is applied to both the position and velocity of a sheet when the Weber number is less than one, we observe absolutely unstable sinuous waves which grow like t1/3. If only the position is perturbed, we find that the sheet is stable and decays like t−2/3 at the origin. Furthermore, if both the position and velocity of a sheet are perturbed in the absence of ambient gas, we observe a new phenomenon in which sinuous waves neither grow nor decay and varicose waves grow like t1/2 with a convective instability.

We consider the buoyancy-driven flow of a fluid injected into a semi-infinite porous medium bounded by a horizontal impermeable barrier through which a single localized sink allows leakage of the injected fluid. Our study is motivated by the geological sequestration of carbon dioxide (CO2), which is less dense than the ambient water, and the possibility that fissures in the bounding ‘cap’ rock may therefore compromise the long-term storage of CO2. A theoretical model is presented in which the leakage through the sink, or fissure, is driven by the hydrostatic pressure at the sink of the injected buoyant fluid. We determine numerical solutions for the evolution of the gravity current in the porous medium and for the quantity of fluid that escapes through the sink as a function of time. A quantity of considerable interest is the efficiency of storage, which we define as the flux of fluid that is stably stored relative to the amount injected. At the later stages in the evolution of the current, the region near the source and sink reaches a quasi-steady state. We find analytical solutions to this asymptotic state which show that the efficiency of storage decreases to zero like 1/lnt, where t is the time since initiation of the current, and predict a dependence on the properties of the sink in agreement with our numerical results. The implications of this result for the geological sequestration of CO2 are discussed.

We consider the propagation of a buoyancy-driven gravity current in a porous medium bounded by a horizontal, impermeable boundary. The current is fed by a constant flux injected at a point and leaks through a line sink at a distance from the injection point. This is an idealized model of how a fault in a cap rock might compromise the geological sequestration of carbon dioxide. The temporal evolution of the efficiency of storage, defined as the instantaneous ratio of the rate at which fluid is stored without leaking to the rate at which it is injected, is of particular interest. We show that the ‘efficiency of storage’ decays like t−2/5 for times t that are long compared with the time taken for the current to reach the fault. This algebraic decay is in contrast to the case of leakage through a circular sink (Neufeld et al., J. Fluid Mech., vol. 2010) where the efficiency of storage decays more slowly like 1/lnt. The implications of the predicted decay in the efficiency of storage are discussed in the context of geological sequestration of carbon dioxide. Using parameter values typical of the demonstration project at Sleipner, Norway, we show that the efficiency of storage should remain greater than 90% on a time scale of millennia, provided that there are no significant faults in the cap rock within about 12km of the injection site.

The behaviour of a miniature calorimetric sensor, which is under consideration for catheter-based coronary-artery-flow assessment, is investigated in both steady and pulsatile tube flows. The sensor is composed of a heating element operated at constant power and two thermopiles that measure flow-induced temperature differences over the sensor surface. An analytical sensor model is developed, which includes axial heat conduction in the fluid and a simple representation of the solid wall, assuming a quasi-steady sensor response to the pulsatile flow. To reduce the mathematical problem, described by a two-dimensional advection–diffusion equation, a spectral method is applied. A Fourier transform is then used to solve the resulting set of ordinary differential equations and an analytical expression for the fluid temperature is found. To validate the analytical model, experiments with the sensor mounted in a tube have been performed in steady and pulsatile water flows with various amplitudes and Strouhal numbers. Experimental results are generally in good agreement with theory and show a quasi-steady sensor response in the coronary-flow regime. The model can therefore be used to optimize the sensor design for coronary-flow assessment.

We present an experimental study on the trapped modes occurring around a vertical surface-piercing circular cylinder of radius a placed symmetrically between the parallel walls of a long but finite water waveguide of width 2d. A wavemaker placed near the entrance of the waveguide is used to force an asymmetric perturbation into the guide, and the free-surface deformation field is measured using a global single-shot optical profilometric technique. In this configuration, several values of the aspect ratio a/d were explored for a range of driving frequencies below the waveguide's cutoff. Decomposition of the obtained fields in harmonics of the driving frequency allowed for the isolation of the linear contribution, which was subsequently separated according to the symmetries of the problem. For each of the aspect ratios considered, the spatial structure of the trapped mode was obtained and compared to the theoretical predictions given by a multipole expansion method. The waveguide–obstacle system was further characterized in terms of reflection and transmission coefficients, which led to the construction of resonance curves showing the presence of one or two trapped modes (depending on the value of a/d), a result that is consistent with the theoretical predictions available in the literature. The frequency dependency of the trapped modes with the geometrical parameter a/d was determined from these curves and successfully compared to the theoretical predictions available within the frame of linear wave theory.

A new numerical scheme for obtaining the steady-state form of an internal solitary wave of large amplitude is presented. A stratified inviscid two-dimensional fluid under the Boussinesq approximation flowing between horizontal rigid boundaries is considered. The stratification is stable, and buoyancy is continuously differentiable throughout the domain of the flow. Solutions are obtained by tracing the buoyancy frequency along streamlines from the undisturbed far field. From this the vorticity field can be constructed and the streamfunction may then be obtained by inversion of Laplace's operator. The scheme is presented as an iterative solver, where the inversion of Laplace's operator is performed spectrally. The solutions agree well with previous results for stratification in which the buoyancy frequency is a discontinuous function. The new numerical scheme allows significantly larger amplitude waves to be computed than have been presented before and it is shown that waves with Richardson numbers as low as 0.062 can be computed straightforwardly. The method is also extended to deal in a novel way with closed streamlines when they occur in the domain. The new solutions are tested in independent fully nonlinear time-dependent simulations and are verified to be steady. Waves with regions of recirculation are also discussed.

The vortex formation behind an orifice is a widely investigated phenomenon, which has been recently studied in several problems of biological relevance. In the case of a circular opening, several works in the literature have shown the existence of a limiting process for vortex ring formation that leads to the concept of critical formation time. In the different geometric arrangement of a planar flow, which corresponds to an opening with straight edges, it has been recently outlined that such a concept does not apply. This discrepancy opens the question about the presence of limiting conditions when apertures with irregular shape are considered. In this paper, the three-dimensional vortex formation due to the impulsively started flow through slender openings is studied with the numerical solution of the Navier–Stokes equations, at values of the Reynolds number that allow the comparison with previous two-dimensional findings. The analysis of the three-dimensional results reveals the two-dimensional nature of the early vortex formation phase. During an intermediate phase, the flow evolution appears to be driven by the local curvature of the orifice edge, and the time scale of the phenomena exhibits a surprisingly good agreement with those found in axisymmetric problems with the same curvature. The long-time evolution shows the complete development of the three-dimensional vorticity dynamics, which does not allow the definition of further unifying concepts.

Run-up on a truncated impermeable beach is analysed theoretically and experimentally to find the volume of fluid, associated with a single wave event, that flows over the end of the beach. The theoretical calculations investigate the motion using the shallow-water equations and the fluid is allowed to flow freely over the end of the beach. Two models of wave events are considered: dam-break initial conditions, in which fluid collapses from rest to run-up and overtop the beach, and a waveform that models swash associated with the collapse of a long solitary bore. The calculations are made using quasi-analytical techniques, following the hodograph transformation of the governing equations. They yield predictions for the volume of fluid per unit width that overtops the beach, primarily as a function of the dimensionless length of the beach. These predictions are often far in excess of previous theoretical calculations. New experimental results are also reported in which the overtopping volumes due to flows initiated from dam-break conditions are studied for a range of reservoir lengths and heights and for a range of lengths and inclinations of the beach. Without the need for any empirically fitted parameters, good agreement is found between the experimental measurements and the theoretical predictions in regimes for which the effects of drag are negligible.