It is shown that the pressure signal measured at the outer edge of a jet mixing layer is entirely hydrodynamic in nature and provides a good measure of the large-scale structure of the turbulent flow. Measurement of the pressure signal provides a unique opportunity to utilize proper orthogonal decomposition (POD) to deduce the streamwise structure. Since pressure is a scalar, a significant reduction in the numerical and experimental complexity inherent in the analysis of velocity vector fields results.

The POD streamwise eigenfunctions show that the structure associated with any frequency–azimuthal mode number combination displays the general characteristics of amplification–saturation–decay of an instability wave, all within about three wavelengths. High-frequency components saturate early in x and low-frequency components saturate further downstream, indicative of the inhomogeneous character of the flow in the streamwise direction. Application of the POD technique allows the phase velocity to be determined taking into account the inhomogeneity of the flow in the streamwise direction. The phase velocity of each instability wave (POD eigenvector) is constant and equal to 0.58Uj, indicating that the jet structure is non-dispersive.

Using the shot-noise decomposition, a characteristic event is constructed. This event is found to contain evidence of both pairings and triplings of vortex structures. The tripling results in a rapid increase in the first asymmetric (m=1) component. On average, pairing occurs once every four Uj/D while tripling occurs once every 13Uj/D.

A neutrally buoyant droplet in a fluid possessing a temperature gradient migrates under the action of thermocapillarity. The drop pole in the high-temperature region has a reduced surface tension. The surface pulls away from this low-tension region, establishing a Marangoni stress which propels the droplet into the warmer fluid. Thermocapillary migration is retarded by the adsorption of surfactant: surfactant is swept to the trailing pole by surface convection, establishing a surfactant-induced Marangoni stress resisting the flow (Barton & Subramanian 1990).

The impact of surfactant adsorption on drop thermocapillary motion is studied for two nonlinear adsorption frameworks in the sorption-controlled limit. The Langmuir adsorption framework accounts for the maximum surface concentration Γ′∞ that can be attained for monolayer adsorption; the Frumkin adsorption framework accounts for Γ′∞ and for non-ideal surfactant interactions. The compositional dependence of the surface tension alters both the thermocapillary stress which drives the flow and the surfactant-induced Marangoni stress which retards it. The competition between these stresses determines the terminal velocity U′, which is given by Young's velocity U′0 in the absence of surfactant adsorption. In the regime where adsorption–desorption and surface convection are of the same order, U′ initially decreases with surfactant concentration for the Langmuir model. A minimum is then attained, and U′ subsequently increases slightly with bulk concentration, but remains significantly less than U′0. For cohesive interactions in the Frumkin model, U′ decreases monotonically with surfactant concentration, asymptoting to a value less than the Langmuir velocity. For repulsive interactions, U′ is non-monotonic, initially decreasing with concentration, subsequently increasing for elevated concentrations. The implications of these results for using surfactants to control surface mobilities in thermocapillary migration are discussed.

We consider two-dimensional free surface flow caused by a pressure wavemaker in a viscous incompressible fluid of finite depth and infinite horizontal extent. The governing equations are expressed in dimensionless form, and attention is restricted to the case δ[Lt ]ε[Lt ]1, where δ is the characteristic dimensionless thickness of a Stokes boundary layer and ε is the Strouhal number. Our aim is to provide a global picture of the flow, with emphasis on the steady streaming velocity.

The asymptotic flow structure near the wavenumber is found to consist of five distinct vertical regions: bottom and surface Stokes layers of dimensionless thickness O(δ), bottom and surface Stuart layers of dimensionless thickness O(δ/ε) lying outside the Stokes layers, and an irrotational outer region of dimensionless thickness O(1). Equations describing the flow in all regions are derived, and the lowest-order steady streaming velocity in the near-field outer region is computed analytically.

It is shown that the flow far from the wavemaker is affected by thickening of the Stuart layers on the horizontal length scale O[(ε/δ)2], by viscous wave decay on the scale O(1/δ), and by nonlinear interactions on the scale O(1/ε2). The analysis of the flow in this region is simplified by imposing the restriction δ=O(ε2), so that all three processes take place on the same scale. The far-field flow structure is found to consist of a viscous outer core bounded by Stokes layers at the bottom boundary and water surface. An evolution equation governing the wave amplitude is derived and solved analytically. This solution and near-field matching conditions are employed to calculate the steady flow in the core numerically, and the results are compared with other theories and with observations.

Chaotic behaviour is found for sufficiently long triaxial ellipsoidal non-Brownian particles immersed in steady simple shear flow of a Newtonian fluid in an inertialess approximation. The result is first determined via numerical simulations. An analytic theory explaining the onset of chaotic rotation is then proposed. The chaotic rotation coexists with periodic and quasi-periodic motions. Quasi-periodic motions are depicted by regular closed loops and islands in the system Poincaré map, whereas chaotic rotations form a stochastic layer.

The unsteady inviscid two-dimensional flow field and the wave configurations which result when a supersonic vehicle strikes a planar oblique shock wave were modelled and analytically predicted using some approximations and simplifying assumptions. Based on the two- and three-shock theories together with the geometric shock dynamics theory, both regular (windward) and irregular (leeward) shock-on-shock (S-O-S) interactions were investigated, and the transition criterion between them was suggested. For the case of regular S-O-S interaction, the transmitted shock wave reflects over the vehicle body surface either as a regular (RR) or a Mach reflection (MR) depending on the inclination angle and the strength of the impingement shock wave. A pronounced peak surface pressure jump was found to exist during the transition from RR to MR. A RR[harr ]MR transition criterion when the flow ahead of the shock pattern is not quiescent was proposed. Predictions based on the model developed here are superior to those of approximate theories when compared to the available experimental data and numerical simulations.

The paper describes a theory of particle deposition based formally on the conservation equations of particle mass and momentum. These equations are formulated in an Eulerian coordinate system and are then Reynolds averaged, a procedure which generates a number of turbulence correlations, two of which are of prime importance. One represents ‘turbulent diffusion’ and the other ‘turbophoresis’, a convective drift of particles down gradients of mean-square fluctuating velocity. Turbophoresis is not a small correction; it dominates the particle dynamic behaviour in the diffusion-impaction and inertia-moderated regimes.

Adopting a simple model for the turbophoretic force, the theory is used to calculate deposition from fully developed turbulent pipe flow. Agreement with experimental measurements is good. It is found that the Saffman lift force plays an important role in the inertia-moderated regime but that the effect of gravity on deposition from vertical flows is negligible. The model also predicts an increase in particle concentration close to the wall in the diffusion-impaction regime, a result which is partially corroborated by an independent ‘direct numerical simulation’ study.

The new deposition theory represents a considerable advance in physical understanding over previous free-flight theories. It also offers many avenues for future development, particularly in the simultaneous calculation of laminar (pure inertial) and turbulent particle transport in more complex two- and three-dimensional geometries.

The general purpose of this paper is to investigate some consequences of the randomness of the velocities of interacting rigid particles falling under gravity through viscous fluid at small Reynolds number. Random velocities often imply diffusive transport of the particles, but particle diffusion of the conventional kind exists only when the length characteristic of the diffusion process is small compared with the distance over which the particle concentration is effectively uniform. When this condition is not satisfied, some alternative analytical description of the dispersion process is needed. Here we suppose that a dilute dispersion of sedimenting particles is bounded externally by pure fluid and enquire about the rate at which particles make outward random crossings of the (imaginary) boundary. If the particles are initially distributed with uniform concentration within a spherical boundary, we gain the convenience of approximately steady conditions with a velocity distribution like that in a falling spherical drop of pure liquid. However, randomness of the particle velocities causes some particles to make an outward crossing of the spherical boundary and to be carried round the boundary and thence downstream in a vertical ‘tail’. This is the nature of break-up of a falling cloud of particles.

A numerical simulation of the motion of a number of interacting particles (maximum 320) assumed to act as Stokeslets confirms the validity of the above picture of the way in which particles leak away from a spherical cluster of particles. A dimensionally correct empirical relation for the rate at which particles are lost from the cluster involves a constant which is indeed found to depend only weakly on the various parameters occurring in the numerical simulation. According to this relation the rate at which particles are lost from the blob is proportional to the fall speed of an isolated particle and to the area of the blob boundary. Some photographs of a leaking tail of particles in figure 5 also provide support for the qualitative picture.

This paper provides a new study of the axisymmetric vortex breakdown phenomenon. Our approach is based on a thorough investigation of the axisymmetric unsteady Euler equations which describe the dynamics of a swirling flow in a finite-length constant-area pipe. We study the stability characteristics as well as the time-asymptotic behaviour of the flow as it relates to the steady-state solutions. The results are established through a rigorous mathematical analysis and provide a solid theoretical understanding of the dynamics of an axisymmetric swirling flow. The stability and steady-state analyses suggest a consistent explanation of the mechanism leading to the axisymmetric vortex breakdown phenomenon in high-Reynolds-number swirling flows in a pipe. It is an evolution from an initial columnar swirling flow to another relatively stable equilibrium state which represents a flow around a separation zone. This evolution is the result of the loss of stability of the base columnar state when the swirl ratio of the incoming flow is near or above the critical level.

As an extension of the authors' work on isotropic vortical turbulence interacting with a shock wave (Lee, Lele & Moin 1993), direct numerical simulation and linear analysis are performed for stronger shock waves to investigate the effects of the upstream shock-normal Mach number (M1). A shock-capturing scheme is developed to accurately simulate the unsteady interaction of turbulence with shock waves. Turbulence kinetic energy is amplified across the shock wave, and this amplification tends to saturate beyond M1 = 3.0. An existing controversy between experiments and theoretical predictions on length scale change is thoroughly investigated through the shock-capturing simulation: most turbulence length scales decrease across the shock, while the dissipation length scale (ρq3/ε) increases slightly for shock waves with M1<1.65. Fluctuations in thermodynamic variables behind the shock wave are nearly isentropic for M1<1.2, and deviate significantly from isentropy for the stronger shock waves, due to the entropy fluctuation generated through the interaction.

An analytical dispersion relation describing the linear stability of a plane detonation wave to low-frequency two-dimensional disturbances with arbitrary wavenumbers is derived using a normal mode approach and a combination of high activation energy and Newtonian limit asymptotics, where the ratio of specific heats γ→1. The reaction chemistry is characterized by one-step Arrhenius kinetics. The analysis assumes a large activation energy in the plane steady-state detonation wave and a characteristic linear disturbance wavelength which is longer than the fire-zone thickness. Newtonian limit asymptotics are employed to obtain a complete analytical description of the disturbance behaviour in the induction zone of the detonation wave. The analytical dispersion relation that is derived depends on the activation energy and exhibits favourable agreement with numerical solutions of the full linear stability problem for low-frequency one- and two-dimensional disturbances, even when the activation energy is only moderate. Moreover, the dispersion relation retains vitally important characteristics of the full problem such as the one-dimensional stability of the detonation wave to low-frequency disturbances for decreasing activation energies or increasing overdrives. When two-dimensional oscillatory disturbances are considered, the analytical dispersion relation predicts a monotonic increase in the disturbance growth rate with increasing wavenumber, until a maximum growth rate is reached at a finite wavenumber. Subsequently the growth rate decays with further increases in wavenumber until the detonation becomes stable to the two-dimensional disturbance. In addition, through a new detailed analysis of the behaviour of the perturbations near the fire front, the present analysis is found to be equally valid for detonation waves travelling at the Chapman–Jouguet velocity and for detonation waves which are overdriven. It is found that in contrast to the standard imposition of a radiation or piston condition on acoustic disturbances in the equilibrium zone for overdriven waves, a compatibility condition on the perturbation jump conditions across the fire zone must be satisfied for detonation waves propagating at the Chapman–Jouguet detonation velocity. An insight into the physical mechanisms of the one- and two-dimensional linear instability is also gained, and is found to involve an intricate coupling of acoustic and entropy wave propagation within the detonation wave.

Core–annular flow of two fluids is examined at the onset of a non-axisymmetric instability. This is a pattern selection problem: the bifurcating solutions are travelling waves and standing waves. The former travel in the azimuthal direction as well as the axial direction and would be observed as corkscrew waves. The standing waves travel in the axial direction but not in the azimuthal direction and appear as snakes. Weakly nonlinear interactions are studied to see whether one of these waves will be stable to small-amplitude perturbations. Sample situations for down-flow are discussed. The corkscrews tend to be preferred when the annulus is narrow, while snakes are more likely when the annulus is wide.

The velocity distribution function for the steady shear flow of disks (in two dimensions) and spheres (in three dimensions) in a channel is determined in the limit where the frequency of particle–wall collisions is large compared to particle–particle collisions. An asymptotic analysis is used in the small parameter ε, which is naL in two dimensions and n2L in three dimensions, where n is the number density of particles (per unit area in two dimensions and per unit volume in three dimensions), L is the separation of the walls of the channel and a is the particle diameter. The particle–wall collisions are inelastic, and are described by simple relations which involve coefficients of restitution et and en in the tangential and normal directions, and both elastic and inelastic binary collisions between particles are considered. In the absence of binary collisions between particles, it is found that the particle velocities converge to two constant values (ux, uy) =(±V, 0) after repeated collisions with the wall, where ux and uy are the velocities tangential and normal to the wall, V=(1−et) Vw/(1+et), and Vw and −Vw are the tangential velocities of the walls of the channel. The effect of binary collisions is included using a self-consistent calculation, and the distribution function is determined using the condition that the net collisional flux of particles at any point in velocity space is zero at steady state. Certain approximations are made regarding the velocities of particles undergoing binary collisions in order to obtain analytical results for the distribution function, and these approximations are justified analytically by showing that the error incurred decreases proportional to ε1/2 in the limit ε→0. A numerical calculation of the mean square of the difference between the exact flux and the approximate flux confirms that the error decreases proportional to ε1/2 in the limit ε→0. The moments of the velocity distribution function are evaluated, and it is found that 〈u2x〉→V2, 〈u2y〉 ∼V2ε and − 〈uxuy〉 ∼ V2εlog(ε−1) in the limit ε→0. It is found that the distribution function and the scaling laws for the velocity moments are similar for both two- and three-dimensional systems.

A vibrating-element density meter is a mechanical oscillator with known properties, for example a tuning fork or a simple rod, driven to vibrate at a known frequency. The oscillator is immersed in a fluid and the resonant frequency measured. The density of the fluid can then be inferred. We consider an idealized meter immersed in two-phase flows of various types, and investigate whether a simple single-phase interpretation allows us to deduce the density of the mixture. We find that, when the density contrast between the two fluids is not great, the simple interpretation gives good results, for example in oil/water flows. However, when the density contrast is significant, for example in gas/liquid flows, the simple interpretation is highly inaccurate.

This paper is concerned with the nonlinear interaction and development of a pair of oblique Tollmien–Schlichting waves which travel with equal but opposite angles to the free stream in a boundary layer. Our approach is based on high-Reynolds-number asymptotic methods. The so-called ‘upper-branch’ scaling is adopted so that there exists a well-defined critical layer, i.e. a thin region surrounding the level at which the basic flow velocity equals the phase velocity of the waves. We show that following the initial linear growth, the disturbance evolves through several distinct nonlinear stages. In the first of these, nonlinearity only affects the phase angle of the amplitude of the disturbance, causing rapid wavelength shortening, while the modulus of the amplitude still grows exponentially as in the linear regime. The second stage starts when the wavelength shortening produces a back reaction on the development of the modulus. The phase angle and the modulus then evolve on different spatial scales, and are governed by two coupled nonlinear equations. The solution to these equations develops a singularity at a finite distance downstream. As a result, the disturbance enters the third stage in which it evolves over a faster spatial scale, and the critical layer becomes both non-equilibrium and viscous in nature, in contrast to the two previous stages, where the critical layer is in equilibrium and purely viscosity dominated. In this stage, the development is governed by an amplitude equation with the same nonlinear term as that derived by Wu, Lee & Cowley (1993) for the interaction between a pair of Rayleigh waves. The solution develops a new singularity, leading to the fourth stage where the flow is governed by the fully nonlinear three-dimensional inviscid triple-deck equations. It is suggested that the stages of evolution revealed here may characterize the so-called ‘oblique breakdown’ in a boundary layer. A discussion of the extension of the analysis to include the resonant-triad interaction is given.

The evolution of a wavepacket in a laminar boundary layer is studied experimentally, paying particular attention to the stage just prior to the formation of a turbulent spot. The initial stages of development are found to be in very good agreement with previous results and indicate a stage in which the disturbance grows according to linear theory followed by a weakly nonlinear stage in which the subharmonic grows, apparently through a parametric resonance mechanism. In a third stage, strong non-linear interactions are observed in which the disturbance develops a streaky structure and the corresponding wavenumber–frequency spectra exhibit an organized cascade mechanism in which spectral peaks appear with increasing spanwise wavenumber and with frequencies which alternate between zero and the subharmonic frequency. Higher harmonics are also observed, although with lower amplitude than the low-frequency peaks. The final (breakdown) stage is characterized by the appearance of high-frequency oscillations with random phase, located at low-speed ‘spike’ regions of the primary disturbance. Wavelet transforms are used to analyse the structure of both coherent and random small-scale structure of the disturbance. In particular, the breakdown oscillations are also observed to have a wavepacket character riding on the large-amplitude primary disturbance.