Making a normal-mode assumption, we shall investigate stability with respect to non-axisymmetric perturbations of an inhomogeneous incompressible fluid rotating between two perfectly conducting, infinite, coaxial cylinders in the presence of an axial and a toroidal magnetic field. We shall establish sufficient conditions for stability, discuss the westward-drift nature of unstable modes and estimate upper bounds on the azimuthal phase speeds and growth rates. There will also be a discussion of the validity of the sufficient conditions for stability when the normal-mode assumption is not made, so that the stability is based on an initial-value problem.

sfjsdjfsdfsdfjsdjfsf

The flows caused by a uniformly expanding circular cylinder or sphere in a perfect gas at rest at various (from small to very large) expansion velocities are analysed by a method of successive approximations in which the first approximation represents the incompressible flow. The shock waves which form around the bodies are also treated. The results for a sphere, even up to the third approximation, agree closely with Taylor's (1946) calculations for all the expansion rates.

If the no-slip condition is used to determine the flow produced when a fluid interface moves along a solid boundary, a non-integrable stress is obtained. In part 1 of this study (Hocking 1976), it was argued that, when allowance was made for the presence of irregularities on the solid boundary, an effective slip coefficient could be found, which might remove the difficulty.

This paper examines the effect of a slip coefficient on the flow in the neighbourhood of the contact line. Particular cases which are solved in detail are liquid–gas interfaces at an arbitrary angle, and normal contact of fluids of arbitrary viscosity. The contribution of the vicinity of the contact line to the force on the boundary is obtained.

The inner region, near the contact line, must be matched with an outer flow, in which the no-slip condition can be applied, in order to obtain the total value of the force on the boundary. This force is determined for the flow of two fluids between parallel plates and in a pipe, with a plane interface. The enhanced resistance produced by the presence of the interface is calculated, and it is shown to be equivalent to an increase in the length of the column of fluid by a small multiple of the pipe radius.

The manner in which fluid driven through a channel of width a responds in anticipation of a severe asymmetric distortion (e.g. to the wall or interior conditions downstream) is discussed when the oncoming flow is fully developed, the characteristic Reynolds number K is large and the whole motion remains laminar. Far ahead of the disturbance, at distances $O(aK^{\frac{1}{7}})$, there occurs a free interaction which triggers a small displacement in the core, generating viscous layers near both walls; the relatively large induced pressure gradient acting across the channel is then found to sustain the growth of this displacement. Numerical solutions of the fundamental nonlinear problem show that one layer separates in a regular fashion, but that beyond separation the other layer, under compression, produces a singularity in the interaction. Analysis of the singularity based on a self-similar structure in the partly reversed flow then leads to a description of the nonlinear flow features nearer the distortion which seems to have strong physical significance. The major implication is that the flow will have already separated at one wall, and developed a definite nonlinear character quite distinct from that of the original Poiseuille flow, long before it reaches finite distances from the distortion. The separation point is predicted to be a distance $0.49 aK^{\frac{1}{7}} (+ O(a))$ ahead of the particular finite distortion. Comparisons of this and other predictions with computed solutions of the full Navier–Stokes equations show reasonable agreement.

Lighthill, in his elegant and classic theory of jet noise, showed that the far-field acoustic pressure of noise generated by turbulence is proportional to the integral over the jet volume of the second time derivative of the Lighthill stress tensor, the integrand being evaluated at a retarded time. The purpose of this paper is to generalize the above results to include the effects of mean flow (velocity and temperature) surrounding the source of sound. It is shown quite generally that the integrand is now a certain functional of the Lighthill stress tensor evaluated at a retarded time. More important, however, at low and high frequencies this functional assumes an extremely simple form, so that the acoustic field can once more be given by integrals of the time derivatives of the Lighthill tensor. Both the self- and the shear-noise contributions to the pressure are evaluated.

The motion of a continuously stratified fluid in a rotating annulus is investigated experimentally. The flow is driven by the differential rotation of a lid on the annulus in contact with the fluid. For small enough Rossby numbers the basic flow is found to be axisymmetric, but in contrast to the unstratified case the zonal flow is not independent of the depth, and there is significant vertical shear in the interior. As the Rossby number of the flow is increased the axisymmetric flow becomes unstable to non-axisymmetric baroclinic disturbances. Some features of these disturbances as well as the time response to the motion of the lid are examined.

A visualization method is used to obtain the main features of the hydrodynamic field for flow past a circular cylinder moving at a uniform speed in a direction perpendicular to its generating lines in a tank filled with a viscous liquid. Photographs are presented to show the particular fineness of the experimental technique. More especially, the closed wake and the velocity distribution behind the obstacle are investigated; the changes in the geometrical parameters describing the eddies with Reynolds number (5 < Re < 40) and with the ratio λ between the diameters of the cylinder and tank are given. A comparison with existing numerical and experimental results is presented and some remarks are made about the calculation techniques proposed up to the present. The limits of the Reynolds-number range for which the twin vortices exist and adhere stably to the cylinder are determined.

The plane flow induced by the impulsive start of a circular cylinder previously at rest in a still fluid is investigated experimentally by a visualization technique. The details of the flow field at the different stages of its establishment are pointed out, and the effect of the wall upon the evolution of the flow in time is examined. Photographs of the flow patterns are presented. This study corresponds to that range of Reynolds numbers for which a closed wake exists and adheres stably to the cylinder.

This paper investigates the propulsive performance of the lunate tails of aquatic animals achieving high propulsive efficiency (the hydromechanical efficiency being defined as the ratio of the work done by the mean forward thrust to the mean rate at which work is done by the tail movements on the surrounding fluid). Small amplitude heaving and pitching motions of a finite flat-plate wing of general planform with a rounded leading edge and a sharp trailing edge are considered. This is a generalization of Chopra's (1974) work on model rectangular tails. This motion characterizes vertical oscillations of the horizontal tail flukes of some cetacean mammals. The same oscillations, turned through a right angle to become horizontal motions of side-slip and yaw, characterize the caudal fins of certain fast-swimming fishes; viz. wahoo, tunny, wavyback skipjack, etc., from the Percomorphi and whale shark, porbeagle, etc., from the Selachii. Davies’ (1963, 1976) method of finding the loading distribution on the wing and generalized force coefficients, through approximate solution of an integral equation relating the loading and the upwash (lifting-surface theory), is used to find the total thrust and the rate of working of the tail, which in turn specify the hydromechanical swimming performance of the animals. The physical parameters concerned are the tail aspect ratio ((span)2/planform area), the reduced frequency (angular frequency x typical length/forward speed), the feathering parameter (the ratio of the tail slope to the slope of the path of the pitching axis), the position of the pitching axis, and the curved shapes of the leading and trailing edges. The variation of the thrust and the propulsive efficiency with these parameters has been discussed to indicate the optimum shape of the tail. It is found that, compared with a rectangular tail, a curved leading edge as in lunate tails gives a reduced thrust contribution from the leading-edge suction for the same total thrust; however, a sweep angle of the leading edge exceeding about 30° leads to a marked reduction of efficiency. Another implication of the present analysis is that no negative work is involved in the actual oscillation of the tail.

The present results are used to obtain an estimate of the drag coefficient for the motion of the animals, based on observed data and the computed thrust. The results show some evidence of differences between the CD's for cetacean mammals and scombroid fish respectively. Some discussion of this difference is also given.

We consider the flow of rapidly rotating fluid over topography in a circular basin. The equations of motion (here the inviscid quasi-geostrophic vorticity equations) can be integrated exactly for certain zonally averaged currents. The assumption of the existence of a specified zonal current is equivalent to the assumption of no upstream influence in the unbounded case. It is unlikely that such solutions can be realized in experiments with real fluids for the presence of viscosity, however small, causes ‘zonal influence’ independent of the magnitude of the viscosity at times larger than the spin-up time. For times smaller than the spin-up time decaying transients can cause zonal influence which increases in magnitude with decreasing viscosity.

This is a review of an IUTAM symposium held in Canberra on 20–23 July 1976. The subject matter ranged widely: from surf on beaches to waves of oceanic scale, and from elegant analysis to observations from tide gauges.

The growth of the unsteady boundary layer on an infinite rotating disk in a counter-rotating fluid is examined numerically and analytically. The numerical computations indicate that the boundary layer breaks down when ωt* ≈ 2·36 in a novel way: the displacement thickness, as well as all the velocity components, becomes infinite. This numerical solution is fitted to an asymptotic expansion which contains the singularities found in the numerical integrations, and it is concluded that the solution of the unsteady similarity equations does break down at a finite time as the numerical results indicate. This problem is placed in a physically more realistic context by considering numerically the unsteady boundary layer which develops on a finite rotating disk in a counter-rotating fluid. It is found that the breakdown of the solution occurs at the axis at the same time, and thus the concept of a thin boundary layer in this more realistic problem is also destroyed in a finite time.

Natural modes of water oscillation inside harbours are known to occur with periods of the order of minutes. It seems likely that these oscillations are excited by water fluctuations of similar period outside the harbour and an often quoted cause of such fluctuations is the phenomenon of surf beats. These are thought to be long waves which are reflected back out to sea when a primary wave system breaks upon a beach. In this paper it is shown theoretically that the natural oscillations of a harbour can be excited directly, without breaking of the primary wave system, by set-down beneath wave groups, which is a long-period disturbance travelling towards the shore line at the group velocity. This theory is in agreement with model experimental results which show that, when the group period is close to a natural period of the harbour, resonance will occur with the set-down behaving as if it were a real long wave.

The slope and curvature distributions of wind waves along two principal axes (upwind-downwind and cross-wind) have been measured in a laboratory tank under various wind velocities. In both directions, the slope distributions are very closely Gaussian, and the components of the mean-square water-surface slope vary loga,rithmically with the friction velocity of the wind. As the windvelocityincreases, the ratio of the upwind-downwind and cross-wind components increases and lies between 0.5 and 0.6 at high wind velocities in the gravity-governed regime of wind-wave interaction. The radius of .water-surface curvature, along either direction of measurement, is generally found to be greater at a steeper viewing angle from the normal to the mean water surface. The average radius of curvature of the disturbed surface varies inversely with the friction velocity of the wind. The ratio of the upwind-downwind and cross-wind components of the average radius of curvature is unity at all wind velocities, indicating that the wind-disturbed water surface is isotropic on the smallest scale. Other results show that both the slope and the curvature distributions are asymmetric along the upwind-downwind direction, either because of the presence of parasitic capillaries or because of the occurrence of wave breaking. The results also indicate that even the high frequency portion of the spectrum is saturated locally but the spectrum is not universal, and that the long waves suppress the growth of the nearly saturated ripples.

The excitation of standing edge waves of frequency ½ω by a normally incident wave train of frequency ω has been discussed previously (Guza & Davis 1974; Guza & Inman 1975; Guza & Bowen 1976) on the basis of shallow-water theory. Here the problem is formulated in the full water-wave theory without making the shallow-water approximation and solved for beach angles β = π/2N, where N is an integer. The work confirms the shallow-water results in the limit N [Gt ] 1, shows the effect of larger beach angles and allows a more complete discussion of some aspects of the problem.

The wave-mechanical condition (Landahl 1972) for breakdown of an unsteady laminar flow into strong small-scale secondary instabilities is applied to some simple stratified inviscid shear flows. The cases considered have one or two discrete density interfaces and simple discontinuous or continuous velocity profiles. A primary wavelike disturbance to such a flow produces a perturbation velocity that is discontinuous at a density interface. The resulting instantaneous system, defined as the sum of the mean flow and the primary oscillation, develops a local secondary shear-flow instability that has a group velocity equal to the arithmetic mean of the instantaneous velocities on the two sides of the interface. According to the breakdown criterion, the disturbed flow will become critical whenever this velocity reaches a value equal to the phase velocity of the primary wave. The calculations show that for a single density interface breakdown may occur for low to moderate wave amplitudes in a fairly narrow range of Richardson numbers on the stable side of the stability boundary. On the other hand, in the unstable regime maximum wave slopes of order unity may be reached before breakdown occurs; this conclusion is in qualitative agreement with experiments. When the system includes two density interfaces, it is found that there exists a range of high Richardson numbers far into the stable regime for which breakdown may take place even for very small and zero wave interface deflexions.

Numerical calculations of the Landau constant are presented for the case of a shear layer of finite Reynolds number, having the velocity profile $\overline{u} = \tan h\,y$. It is found that this parameter has a strong dependence on the Reynolds number for Re [les ] 100. In particular, the Landau constant is reduced by 43% from its inviscid value when Re = 40, the latter value being typical of many experiments. This percentage, however, is based upon a calculation in which the mean-flow distortion has been neglected. A rough estimate of the latter effect indicates that it could possibly increase the value of the Landau constant sufficiently that the net influence of a finite Reynolds number would be of a smaller magnitude.

A double scale technique is used to determine the asymptotic behaviour of a rolled-up vortex sheet. The technique relies on a process of averaging out the saw-tooth-like behaviour of the flow variables, which generates a continuous solution having the structure of a vortex filament. The fine-scale behaviour of the flow is described and includes concentrated vorticity on the sheet. Application to the conical vortex sheet allows the solution of Mangler & Weber (1967) to be rederived. A further application, to Kaden's problem, is worked out and the results are in complete agreement with Moore's asymptotic formulae for the shape of the spiral.

This study contributes to the theory of amplitude vacillation for finite amplitude baroclinic waves in a two-layer, quasi-geostrophic, zonal flow as worked out by Pedlosky. In a recent paper the author has shown that Pedlosky omitted a certain side-wall boundary condition on the mean zonal flow. The neglect of this boundary condition results in an unspecified energy source at the side-wall boundaries and the physical problem is incorrectly posed.

In this paper, Pedlosky's analysis is repeated but with the side-wall boundary condition included. It is shown that the side-wall energy source is negligible only when the zonal wavenumber of the disturbance is large compared with the meridional wavenumber, and not otherwise. Moreover, the energy conversions to and from mean zonal kinetic energy corresponding to Pedlosky's calculations and those given here have essential differences, although for fixed meridional wavenumber, these differences become less pronounced as the zonal wavenumber increases.

It is also shown that, when the side-wall condition is included, the mean flow distortion associated with the wave is different in structure to that which occurs when the condition is omitted. However, as the total disturbance wavenumber a increases, the influence of the side wall on the mean flow structure is confined to a boundary layer of width comparable with the internal deformation radius 2½a−1.

Even when the deformation radius is comparable with the channel width, the conclusions of Pedlosky (1972) concerning the existence of stable periodic solutions are correct, providing viscous effects are vanishingly small, and the criteria for stability of the steady solutions obtained herein are not significantly different from those given by Pedlosky. In this viscous regime, we have also studied the evolution to limit-cycle solutions.

Evolution in the case where viscous effects are small on the time scale for the initial growth of an incipient wave, but not vanishingly small, will be discussed in a subsequent paper.