The transmission of sound out of a semi-infinite circular jet pipe in the presence of subsonic flow from the pipe is investigated. An unstable cylindrical vortex layer attached to the edge of the pipe is considered across which differences in mean subsonic flow, density and temperature are included. A solution satisfying the Kutta condition and causality is found which possesses an instability wave term that dominates within a region of approximately 45° to the downstream jet axis. It is shown that when an exterior flow is imposed the noise level increases upstream whilst the instability wave weakens downstream. The stable part of the solution is shown to agree very well with some recent experimental results.

The Orr–Sommerfeld equation is solved numerically using expansions in Chebyshev polynomials and the generalized Rayleigh quotient iteration. Accurate results for large values of the parameters are obtained, and these further verify the belief that plane Couette flow is stable to infinitesimal disturbances. For finite disturbances, a formal expansion based on the method of Stuart and Watson as modified by Reynolds & Potter is used. This method shows a transition to instability for a large enough amplitude.

The solution of the steady-state Navier–Stokes equations in three dimensions has been obtained by a numerical method for the problem of natural convection in a rectangular cavity as a result of differential side heating. In the past, this problem has generally been treated as though it were two-dimensional. The solutions explore the three-dimensional motion generated by the presence of no-slip adiabatic end walls. For Ra = 104, the three-dimensional motion is shown to be the result of the inertial interaction of the rotating flow with the stationary walls together with a contribution arising from buoyancy forces generated by longitudinal temperature gradients. The inertial effect is inversely dependent on the Prandtl number, whereas the thermal effect is nearly constant. For higher values of Ra, multiple longitudinal flows develop which are a delicate function of Ra, Pr and the cavity aspect ratios.

Perturbation expansions are derived to second order in a wavenumber parameter for the unsteady lift induced on an aerofoil by disturbances convected past it at subsonic speeds. The results are used to discuss other approximate methods which have been used to predict the unsteady forces and noise generated by an aerofoil in turbulent flow.

Linear sinusoidal disturbances are introduced into the boundary layer over a heated flat plate of uniform surface temperature using a vibrating ribbon. The experiment is performed in a low turbulence water tunnel with free-stream turbulence intensities of 0[sdot ]1–0[sdot ]2%. Measurements are made using temperature-compensated hot-film anemometry. Neutral-stability characteristics obtained for the unheated case agree favourably with previous results obtained both in water and in air. Neutral-stability and spatial disturbance-growth-rate characteristics measured for wall temperatures up to 8°F above the free-stream temperature verify trends established by parallel-flow solutions of the disturbance momentum and energy equations. Disturbance growth rates and the band of amplified disturbance frequencies both decrease as wall heating is increased. The experimentally observed increase in the minimum critical Reynolds number Recmin with increased wall heating agrees with the trend predicted by theory. However, effects of non-parallel flow act to reduce the measured values of Recmin by about 120 units compared with predicted parallel-flow values independent of the level of wall heating.

This paper presents a new experimental time-domain technique for the evaluation of the large-scale structure in a turbulent flow. The technique is demonstrated by hot-wire anemometry for a circular jet flow at a moderate Reynolds number of 104, and the large-scale structure identified is compared successfully with smoke flow-visualization observations. The temporal and spatial relationships of the separate large-scale flow events have been derived, and this information enabled the evaluation of the nonlinear spatial development of the large-scale flow structure.

Mass transport close to the sea bottom is investigated for simple harmonic waves around a body with a small horizontal dimension. For gravity waves it is shown that the mass transport very near the bottom points towards a convex corner, but near the top of the boundary layer its direction reverses. Possible implications for silting near a pile and a harbour entrance are discussed and some experimental evidence given. For tides, the Coriolis force introduces a spiralling variation within the boundary layer, and possible inferences for coastline modification are drawn.

Observations by Cornhill ' Roach (1976) of sudanophilic lesions in the vicinity of intercostal arteries in rabbit aortas have shown that lesions develop on the downstream side of the associated ostia. There is considerable conjecture as to the role which varying levels of wall shear stress play in the development of such lesions; Cornhill ' Roach implicate high wall shear stress levels. We develop a consistent model of steady boundary-layer flow past a side slot assuming that there is Stokes flow in the side slot and that the main body of the boundary layer remains undisturbed. Our results show that increased levels of wall shear stress occur both upstream and downstream of the slot. If the withdrawal of fluid through the side slot is sufficiently great there may be a stagnation point on the downstream side of the slot. The wall shear stress level near the slot depends on both normal and transverse motions at the mouth of the slot. Indeed, very near the slot, on a length scale comparable with the slot width, the wall shear stress level depends only on the transverse motions at the mouth of the slot.

An implicit finite-difference technique employing orthogonal curvilinear co-ordinates is used to solve the Navier–Stokes equations for peristaltic flows in which both the wall-wave curvature and the Reynolds number are finite (§2). The numerical solutions agree closely with experimental flow visualizations. The kinematic characteristics of both extensible and inextensible walls (§3) are found to have a distinct influence on the flow processes only near the wall. Without vorticity, peristaltic flow observed from a reference frame moving with the wave will be equivalent to steady potential flow through a stationary wavy channel of similar geometry (§4). Solutions for steady viscous flow (§5) are obtained from simulation of unsteady flow processes beginning from an initial condition of potential peristaltic flow. For nonlinear flows due to a single peristaltic wave of dilatation, the highest stresses and energy exchange rates (§6) occur along the wall and in two instantaneous stagnation regions in the bolus core. A series of computations for periodic wave trains reveals that increasing the Reynolds number from 2[sdot ]3 to 251 yields a modest augmentation in the ratio of flow rate to Reynolds number but induces a much greater increase in the shear stress (§7.1). The transport effectiveness is markedly reduced for pumping against a mild adverse pressure drop (§7.2). Increasing the wave amplitude will lead to the development of travelling vortices within the core region of the peristaltic flow (§7.3).

Results of an experimental investigation of the evolution of a nonlinear wave train on deep water are reported. The initial stage of evolution is found to be characterized by exponential growth of a modulational instability, as was first discovered by Benjamin ' Feir. At later stages of evolution it is found that the instability does not lead to wave-train disintegration or loss of coherence. Instead, the modulation periodically increases and decreases, and the wave train exhibits the Fermi–Pasta–Ulam recurrence phenomenon. Results of an earlier study of nonlinear wave packets by Yuen ' Lake, in which solutions of the nonlinear Schrödinger equation were shown to provide quantitatively correct descriptions of the properties of nonlinear wave packets, are applied to describe the experimentally observed wave-train phenomena. A comparison between the laboratory data and numerical solutions of the nonlinear Schrödinger equation for the long-time evolution of nonlinear wave trains is given.

The oscillation of the wall shear stress caused by imposing sound on a turbulent boundary-layer flow constitutes a boundary condition for the solution of the acoustic wave equation. The no-slip condition at the wall requires the excitation of a shear wave which is superimposed on the sound wave. The shear wave propagates into the turbulent medium. The wall impedance (shear stress/velocity) of streamwise polarized shear waves has been measured in two different ways, namely (a) by evaluating the phase velocity and the attenuation of a plane sound, wave which propagates in turbulent pipe flow, and (b) by evaluating the resonance frequency and the quality factor of a longitudinally vibrating glass pipe which carries turbulent flow. The results, which were obtained over a wide range of Strouhal numbers, exhibit very good agreement between the two measuring methods. The wall shear stress impedance is strongly affected by the turbulence. This indicates that the turbulent shear stress is modulated by the shear wave. At all measuring conditions, the propagation of the shear wave was confined essentially to the inner portion of the turbulent boundary layer. In principle, two different Strouhal numbers, based on inner and outer variables respectively, describe the dynamics of the Reynolds stress, even in the inner layer (Laufer & Badri Narayanan 1971). However, it turns out that the outer Strouhal number (based on the diameter and the centre-line velocity) has no noticeable effect on the wall shear stress impedance. The dependence of the impedance on the inner Strouhal number (based on the friction velocity and the viscosity) reveals that the shear wave is strongly reflected at the edge of the viscous sublayer. It is concluded that the stress-to-strain ratio at the edge of the viscous sublayer corresponds either to a viscoelastic medium or even to a medium with negative viscosity.

It is now well established that coherent structures exist in turbulent shear flows. It should be possible to recognize these in the turbulence signals and to program a computer to extract and ensemble average the corresponding portions of the signals in order to obtain the characteristics of the structures. In this work only the u-signal patterns are recognized, using several simple criteria; simultaneously, however, the v or w signals as well as uv or uw are also processed. It is found that simple signal shapes describe the turbulence structures on the average. The u-signal pattern consists of a gradual deceleration from a local maximum followed by a strong acceleration. This pattern is found in over 65% of the total sample in the region of high Reynolds-stress production. The v signal is found to be approximately 180° out of phase with the u signal. These signal shapes can be easily associated with the coherent structures that have been observed visually. Their details have been enhanced by quadrant truncating. These results are compared with randomly generated signals processed by the same method.

The propagation of waves in slightly inhomogeneous dispersive media is conveniently described by a geometrical or kinematic theory. In such frameworks the solution of the propagation problem is constructed by (a) deriving a dispersion relation and determining its characteristic lines and (b) solving an equation expressing the conservation of a field invariant like the wave action. This paper is concerned with the implementation of the last step under general field and boundary conditions. The method presented is based on the derivation of a variational system of differential equations for the geodesic elements of the wave front. The elementary cross-section of the wave front is obtained by integration and the principle of conservation of the field invariant directly yields the field amplitude. In addition, suitable jump conditions are derived for treating specular reflexions at solid boundaries. The method is illustrated by specific problems of interest in aeroacoustics.

As an inclined rod sediments in an unbounded viscous fluid it will drift horizontally but will not rotate. When it approaches a vertical wall, the rod rotates and so turns away from the wall. Illustrative experiments and a slender-body theory of this phenomenon are presented. In an incidental study the friction coefficients for an isolated rod are found by numerical solution of the slender-body integral equation. These friction coefficients are compared with the asymptotic results of Batchelor (1970) and the numerical results of Youngren ' Acrivos (1975), who did not make a slender-body approximation.

The problem of the instability of a uniform, nonlinear, deep-water wave train to infinitesimal long-wave perturbations, first studied by Benjamin ' Feir (1967) and Benjamin (1967), is re-examined. It is found that the apparent discrepancy between the experimental and theoretical growth rates of the instability is associated with the experimental generation of waves which do not have the Stokes wave profiles assumed in the theory. Experimental and theoretical results relating the initial wave steepness and the most unstable long-wave perturbation are used to obtain a correction factor, which is found to account for the mismatch in wave forms and which resolves the discrepancy in growth rates. The results illustrate that, when theory is compared with experiments in which the values of certain higher-order (nonlinear) quantities must be deduced from measurements of first-order quantities, great care must be taken to ascertain that the experimental conditions and the theoretical assumptions are indeed compatible to the required order.

Earlier ideas are combined to produce a systematic approach both to forming the bulk equations of motion of a dilute suspension and to calculating the overall hydrodynamic interactions between the suspended particles. Equations governing averaged field quantities are derived by taking ensemble averages of the conservation laws and constitutive relations. The bulk equations thus produced contain a term in which the averaging is performed holding one particle fixed. If now the same prescription is applied to fields averaged with one particle fixed, equations are produced containing a term averaged with two particles fixed, and so on up an infinite hierarchy. The hierarchy can be truncated in an asymptotic analysis for small particle concentrations.

This approach to the mechanics of suspensions is illustrated by applying it to three problems which have already been well studied by different methods. The problems concern the first effects of hydrodynamic interactions on the bulk stress and sedimentation velocity of a free suspension, and on the permeability of a fixed bed. Earlier results are recovered in a new light. Multiparticle effects, which before have occurred as divergent sums, are seen to arise because the suspension described by the averaged equations assumes a viscosity and density different from the solvent, or in the case of the fixed bed because the suspension starts behaving as a porous medium instead of as a Newtonian solvent. A close connexion is thus revealed between the averaged-equation description of the interactions and a self-consistent-field model.

The two-dimensional motion of a stably stratified fluid containing two solutes with different molecular diffusivities in an inclined slot has recently been examined by Chen (1975, hereafter referred to as I). The two solutes have continuous opposing gradients with the slower-diffusing one more dense at the bottom. It is found that, in the steady state, there exists a slow upward flow along the slope driven by the slight buoyancy difference near the wall, not unlike the solution found by Wunsch (1970) and Phillips (1970) for a single solute. For the time-dependent flow resulting from switching on the diffusivities at t = 0, there may be a flow reversal near the wall depending on the relative magnitude of λ and τ (where λ is the ratio of the density gradient and τ−1 is the ratio of the diffusivity of the faster-diffusing solute T to that of the slower-diffusing one S). By examining the distributions of S and T across the slot, it becomes apparent that in cases with flow reversal double-diffusive instability is likely to occur.

In this paper, we examine the stability of time-dependent double-diffusive convection in an inclined slot both analytically and experimentally. The time-dependent perturbation equations are numerically integrated starting with an initial distribution of small random disturbances in the vorticity. The growth or decay of the kinetic energy of the perturbations serves to indicate whether the flow is unstable or stable. The results show that the flow becomes more unstable (a) with increasing λ at a given angle of inclination with respect to the vertical and (b) with increasing angle of inclination at a given value of λ. Experiments were carried out in a 2[sdot ]54 cm wide slot using sugar and salt solutions at angles of inclination of 30°, 45° and 60°. Results obtained confirm the trends predicted by the analysis. Good agreement was obtained between the predicted and the experimental values of the critical wavelength for the case λ = 0[sdot ]7.

The conventional long-wave equations for waves propagating over fluid of variable depth depend for their formal derivation on a Taylor series expansion of the velocity potential about the bottom. The expansion, however, is not possible if the depth is not an analytic function of the horizontal co-ordinates and it is a necessary condition for its rapid convergence that the depth is also slowly varying. We show that if in the case of two-dimensional motions the undisturbed fluid is first mapped conformally onto a uniform strip, before the Taylor expansion is made, the analytic condition is removed and the approximations implied in the lowest-order equations are much improved.

In the limit of infinitesimal waves of very long period, consideration of the form of the error suggests that by modifying the coefficients of the reformulated equation we may find an equation exact for arbitrary depth profiles. We are thus able to calculate the reflexion coefficients for long-period waves incident on a step change in depth and a half-depth barrier. The forms of the coefficients of the exact equation are not simple; however, for these particular cases, comparison with the coefficients of the reformulated long-wave equation suggests that in most cases the latter may be adequate. This opens up the possibility of beginning to study finite amplitude and frequency effects on regions of rapidly varying depth.

The motions of a floating torus oscillating in response to incident waves are analysed under the assumptions that the incident wavelength is comparable with the radius of the body section and small compared with the larger radius of the torus. This problem serves to illustrate certain features of the strip theory for ship motions, but the axisymmetric geometry and absence of body ends greatly simplify the analysis. Matched asymptotic expansions are used, with the inner solution close to the body section composed of suitable radiation and scattering problems for the two-dimensional circular cylinder. Resonant standing-wave modes in the internal basin have a singular effect upon the hydrodynamic forces acting on the body, and its response to incident waves.

In this paper we solve some model problems by the method of matched expansions and deduce some features of Green's function using the reciprocal theorem to demonstrate that scattering by density gradients can enhance the radiation efficiency of quadrupoles to that of dipoles. If a fluid is subjected to body forces close to and on both sides of a density interface, then the source system degenerates to a quadrupole only if the components of the forces perpendicular to the interface are equal in magnitude and opposite in direction and, additionally, the components of the forces per unit fluid density parallel to the interface are equal but opposite. In other cases dipole rather than quadrupole radiation efficiency is achieved. But we also show that, despite published assertions to the contrary, the radiation efficiency is not enhanced further and there is no monopole contribution to the radiated sound.