At high Reynolds numbers, wake flows become more globally unstable when they are confined within a duct or between two flat plates. At Reynolds numbers around 100, however, global analyses suggest that such flows become more stable when confined, while local analyses suggest that they become more unstable. The aim of this paper is to resolve this apparent contradiction by examining a set of obstacle-free wakes. In this theoretical and numerical study, we combine global and local stability analyses of planar wake flows at to determine the effect of confinement. We find that confinement acts in three ways: it modifies the length of the recirculation zone if one exists, it brings the boundary layers closer to the shear layers, and it can make the flow more locally absolutely unstable. Depending on the flow parameters, these effects work with or against each other to destabilize or stabilize the flow. In wake flows at with free-slip boundaries, flows are most globally unstable when the outer flows are 50 % wider than the half-width of the inner flow because the first and third effects work together. In wake flows at with no-slip boundaries, confinement has little overall effect when the flows are weakly confined because the first two effects work against the third. Confinement has a strong stabilizing effect, however, when the flows are strongly confined because all three effects work together. By combining local and global analyses, we have been able to isolate these three effects and resolve the apparent contradictions in previous work.

Velocity data were obtained in the turbulent wake of a circular cylinder with an orthogonal array of sixteen X-wires, eight in the (x, y)-plane and eight in the (x, z)-plane. By applying the phase-plane technique to these data, three types of critical points (where the velocity is zero and the streamline slope is indeterminate) were identified. Of these, foci and saddle points occurred most frequently, although a significant number of nodes was also found. Flow topology and properties associated with these points were obtained in each plane. Saddle-point regions associated with spanwise vortices provide the dominant contribution to the Reynolds shear stress and larger contributions to the normal stresses than focal regions. The topology was found to be in close agreement with that obtained from other methods of detecting features of the organized motion. The inter-relationship between critical points simultaneously identified in the two planes can provide some insight into the three-dimensionality of the organized motion. Foci in the (x, z)-plane correspond, with relatively high probability and almost negligible streamwise separation, to saddle points in the (x, y)-plane and are interpreted in terms of ribs aligned with the diverging separatrix between consecutive spanwise vortex rolls. Foci in the (x, z)-plane which correspond, with relatively weak probability, to foci in the (x, y)-plane seem consistent with a distortion of the vortex rolls in the (y, z)-plane.

The theory of an oscillating, high-aspect-ratio, lifting surface with a curved centreline (Cheng & Murillo 1984) is applied to a performance analysis of lunate-tail swimming propulsion. Thrust, power and propulsive efficiency are calculated for model lunate tails with various combinations of mode shapes and morphological features to ascertain the viability of the proportional-feathering concept, and to determine the influence of sweep and centreline curvature. One of the principal conclusions concerns the interchangeability of the heaving amplitude of the peduncle (identified with the major pitching axis) with the centreline sweep, and its effect on the propulsive efficiency, while maintaining the same thrust. Hydrodynamic reasons are also offered for the apparent preference for the crescent-moon fin shape over the V-shape at moderate sweep angles, and for the large sweep angles often found in V-shaped fins.

The present study is devoted to characterizing the coherent organization of vortical structures, which can be fitted into the paradigm of the hairpin-packet model, in the streamwise–wall-normal plane of a canonical turbulent boundary layer at $Re_{\unicode[STIX]{x1D70F}}=377{-}1093$. Proper orthogonal decomposition (POD) of the planar velocity fields measured via two-dimensional particle image velocimetry, together with a spatio-temporal coherence analysis, shows that the first four leading-order POD modes share both geometric similarity and dynamic coherence and jointly depict the downstream convection of the large-scale Q2/Q4 events, which can be regarded as the low-order imprints of the hairpin packets. A simple low-order indicator is then proposed to extract the inclined interfaces of the hairpin packets, based on which a two-point conditional correlation analysis forms a statistical picture of the spatial organization of multiple prograde vortices aligned along the interface within one packet. A saturation of the self-similar growth of the streamwise gap between two neighbouring vortices is seen. This implies a detachment of the hairpin packets from the inner layer. Both the detachment height and the saturated streamwise spacing are found to scale as $Re_{\unicode[STIX]{x1D70F}}^{1/2}$.

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.

There is a continuing debate about the role of linear instability waves in the prediction of jet noise. Parallel mean flow models, such as the one proposed by Lilley, usually neglect these waves because they cause the solution to become infinite. The result is then non-causal and can, therefore, be quite different from the more realistic causal solution, especially for the chaotic flows being considered here. The present paper solves the relevant acoustic equations for a non-parallel mean flow by using a vector Green's function approach and requiring that the mean flow be weakly non-parallel, i.e. requiring that the spread rate be small. It demonstrates that linear instability waves must be accounted for in order to construct a proper causal solution to the jet noise problem. Recent experimental results (e.g. see Tam, Golebiowski & Seiner 1996) show that the small-angle spectra radiated by supersonic jets are significantly different from those radiated at larger angles (say, at 90°) and even exhibit dissimilar frequency scalings (i.e. they scale with Helmholtz number as opposed to Strouhal number). The present solution is (among other things) able to explain this rather puzzling experimental result.

It is observed that an interface is established in any open-ended vertical cylinder when a plume flows in the direction of a mean flow through the cylinder. The location of the interface defines the volume flux of the plume because the total discharge in the environment has been entrained at this level. Experiments have been conducted to verify this relation and to define the limits within which the interface is established. It is found that, as the buoyancy in the plume is reduced or the area of the environment is reduced, limits do exist. An application of the interface in industrial-building ventilation is described.

The interaction between shear and buoyancy effects for Bénard convection in plane Couette flow is studied by performing direct numerical simulations. At moderate Rayleigh number (≈10000−50000), shear tends to organize the flow into quasi-two-dimensional rolls parallel to the mean flow and can enhance heat transfer, while at higher Rayleigh number (>150000), shear tends to disrupt the formation of convective plumes and can reduce heat transfer. A significant temporal oscillation in the local Nusselt number was consistently observed at high Rayleigh numbers, a factor that may contribute to the scatter seen in experimental data. This effect, plus the time-varying reversal of the mean temperature gradient in the middle of the channel, is consistent with a flow model in which the dynamics of large-scale, quasi-two-dimensional, counter-rotating vortical cells are alternately driven by buoyancy and inertial effects. An analysis of the energy balance in the flow shows that the conservative pressure diffusion term, which has been frequently neglected in turbulence models, plays a very important dynamical role in the flow evolution and should be more carefully modelled. Most of the turbulent energy production due to mean shear is generated in the boundary layers, while the buoyant production occurs mainly in the relatively uniform convective core. The simulations and the laboratory experiments of Deardorff & Willis (1967) are in very reasonable qualitative agreement, suggesting that the basic dynamics of the flow are being accurately simulated.

A Kolmogorov-type similarity theory of locally homogeneous and isotropic turbulence generated by a Smagorinsky-type large-eddy simulation (LES) at very large LES Reynolds numbers is developed and discussed. The underlying concept is that the LES equations may be considered equations of motion of specific hypothetical fully turbulent non-Newtonian fluids, called ‘LES fluids’. It is shown that the length scale ls = csδ, which scales the magnitude of the variable viscosity in a Smagorinsky-type LES, is the ‘Smagorinsky-fluid’ counterpart of Kolmogorov's dissipation length $\eta = v^{3/4}\epsilon^{-1/4}$ for a Newtonian fluid where ν is the kinematic viscosity and ε is the energy dissipation rate. While in a Newtonian fluid the viscosity is a material parameter and the length ν depends on ε, in a Smagorinsky fluid the length ls is a material parameter and the viscosity depends on ε. The Smagorinsky coefficient cs may be considered the reciprocal of a ‘microstructure Knudsen number’ of a Smagorinsky fluid. A combination of Lilly's (1967) cut-off model with two well-known spectral models for dissipation-range turbulence (Heisenberg 1948; Pao 1965) leads to models for the LES-generated Kolmogorov coefficient αLES as a function of cs. Both models predict an intrinsic overestimation of αLES for finite values of cs. For cs = 0.2 Heisenberg's and Pao's models provide αLES = 1.74 (16% overestimation) and αLES = 2.14 (43% overestimation), respectively, if limcs → (αLES) = 1.5 is ad hoc assumed. The predicted overestimation becomes negligible beyond about cs = 0.5. The requirement cs > 0.5 is equivalent to Δ < 2ls. A similar requirement, L < 2η where L is the wire length of hot-wire anemometers, has been recommended by experimentalists. The value of limcs → (αLES) for a Smagorinsky-type LES at very large LES Reynolds numbers is not predicted by the models and remains unknown. Two critical values of cs are identified. The first critical cs is Lilly's (1967) value, which indicates the cs below which finite-difference-approximation errors become important; the second critical cs is the value beyond which the Reynolds number similarity is violated.

Derivatives of velocity signals obtained in a turbulent boundary layer are examined for correspondence to the lognormal distribution. It is found that there is rough agreement but that unlikely events at high values are much less common in the observed fields than would be inferred from the lognormal distribution. The actual distributions correspond more to those obtained from a random walk with a limited number of steps, so the difference between these distributions and the lognormal may be related to the fact that the Reynolds number is finite.

The third-order structure function is examined, and found to be roughly consistent with the existence of an inertial subrange of a Kolmogoroff equilibrium reacute;gime over a range of scale which is a priori reasonable but which is far less extensive than the $-\frac{5}{3}$ region of the spectrum.

Experiments were performed to investigate some aspects of turbulence in rotating and non-rotating fluid systems where the turbulence was induced by a horizontal grid oscillating vertically. An earlier theory by the second author made use of a planar source of energy, which appeared to be similar to the energy source of the grid, in determining the characteristics of the turbulence at points some distance away. The simplicity of the theory was in the parameterization of the grid ‘action’ by a single quantity K, with dimensions and characteristics of eddy viscosity.

The experimental results provide additional confirmation of the theory in the non-rotating case, and indicate the usefulness of the idealized energy source in the rotating case. In the latter, we measured the propagation of the front separating disturbed and undisturbed fluid, moving along the axis of rotation. The thickness d(t) of the disturbed region increases at first as (Kt)½, as in a non-rotating fluid, until the Rossby number K/Ωd2k becomes of order unity.

Beyond this the disturbances are wavelike and rotationally dominated, and the thickness now increases linearly with time, yielding a speed of propagation for the front proportional to the wave speed (KΩ)½. Finally, the disturbances reach the bottom and the vessel is in statistical steady state. Then a region of thickness dk independent of time is found, and it contains motion that resembles ordinary, three-dimensional turbulence. dk ∼ (K/Ω)½ is analogous to the depth of the turbulent Ekman layer H ∼ (K/Ω)½, where K is taken as an eddy viscosity.

McEwan constructed a similar rotating experiment, although with a different energy source, and observed vortices parallel to the axis of rotation, provided that the Rossby number was less than a critical value. Our observations and theory indicate that the disappearance of the vortices corresponds to h < dk, where h is the total depth of the fluid. At that point, the whole tank is filled with three-dimensional turbulence.

The two-dimensional prediction method of Bradshaw, Ferriss & Atwell (1967), which was based on the empirical conversion of the turbulent energy equation into a ‘transport’ equation for shear stress, is extended to three-dimensional flows satisfying the boundary-layer approximation (which excludes flows near bluff obstacles or streamwise corners). Predictions, using exactly the same empirical data as in two-dimensional flow, agree to within the likely experimental error with a variety of experiments on ‘infinite’ swept wings.

Turbulence measurements under the influence of a transverse magnetic field have been made at Purdue University's Magneto-Fluid-Mechanic Laboratory in a high aspect ratio channel. The Reynolds number range covered was 25000 ≤ Re 282000; the geometry and experimental conditions were such that the experiment approximated turbulent Hartmann flow. The aspect ratio of the channel was 5·8:1, its walls were electrically insulated and the working fluid was mercury. Measurements in the presence of a magnetic field were made of the skin friction coefficient, the mean velocity profiles, the turbulence intensity profiles (both u’ and v’) and the Reynolds stress profiles.

A sudden change in the damping of the Reynolds stresses was manifested by a ‘hump’ in the curves of Cf versus M/Re taken with the Reynolds number held constant. This ‘hump’ occurs as a gentle rise and sudden drop to the Hartmann laminar line of the Cf data. Close examination of the $\overline{u^{\prime}v^{\prime}}$ data near the wall confirms this behaviour, indicating that the turbulent contribution to the shear stress is the controlling factor in this behaviour of Cf. The Reynolds stresses were completely suppressed to zero at high values of the magnetic field, though the turbulence intensities of u’ and v’ were not. The Reynolds stress data are fundamental in revealing the mechanisms which are at work during the suppression of turbulence by a magnetic field.

It was also found that at high magnetic fields, when most of the turbulence was damped, the skin friction coefficient fell below the values predicted by Hartmann's (1937) laminar solution for high values of M/Re. This result was linked to the presence of ‘M-shaped’ velocity profiles in the direction perpendicular to both the magnetic field and the mean velocity vector. The presence of ‘M-shaped’ profiles has not previously been linked to a reduction in Cf.

The self-similar one-dimensional propagation of a strong shock wave in a medium with exponentially varying density and ray-tube area is studied, using the Eulerian approach of Sedov. Conservation integrals analogous to Sedov's are obtained, with the expression for the Lagrangian variable. Calculated results are compared with the predictions of the CCW (Chisnell, Chester and Whitham) approximation. It was found that, in contrast to the implosion case, the propagation parameter from the CCW approximation is in error by 15% or more.

In this paper some approximate results are presented for the problem of the stability of plane Poiseuille flow of a dusty gas, following a formulation given recently by Saffman (1962). It is assumed that the mass concentration of the dust, f, is small, and results are obtained by making a perturbation of the curve of neutral stability for a clean gas, using the approximate solutions given by Stuart (1954). The perturbation equation is expressed in terms of integrals by introducing the adjoint wave function, the calculation of which is described. The integral coefficients were evaluated by numerical integration using a Mercury computer, and the results are illustrated for f = 0.05 by a set of perturbed neutral stability curves at different values of the time relaxation parameter SR varying from 0 to 500. These results, whilst not of great numerical accuracy, are sufficient to show qualitatively how the curve of neutral stability is modified by the presence of the dust.

We calculate the average swimming velocity and dispersion rate characterizing the transport of swimming gyrotactic micro-organisms suspended in homogeneous (simple) shear. These are requisite effective phenomenological coefficients for the macroscale continuum modelling of bioconvection and related collective-dynamics phenomena. The swimming cells are modelled as rigid axisymmetric dipolar particles subject to stochastic Brownian rotations. Calculations are effected via application of the generalized Taylor dispersion scheme. Attention is focused on finite (as opposed to weak) shear. Results indicate that the largest transverse average swimming velocities (essential to gyrotactic focusing) appear shortly after transition from the ‘tumbling’ mode of motion to cells swimming in the equilibrium direction. At sufficiently large shear rates, dispersivity is not monotonically decreasing with external-field intensity. Exceptional dispersion rates which are unique to non-spherical cells appear in the ‘intermediate domain’ of external fields. These are rationalized in terms of the corresponding deterministic problem (i.e. in the absence of diffusion) when cell rotary motion is governed by the simultaneous coexistence of multiple stable attractors.

An experimental investigation was made of the stability of a two-dimensional jet at low Reynolds numbers with extremely small residual disturbances both in and around the jet. The velocity distribution of a laminar jet is in agreement with Bickley's theoretical result. The stability and transition of a laminar jet are characterized by the Reynolds number based on the slit width and the maximum velocity of the jet. When the Reynolds number is less than 10, the whole jet is laminar. When the Reynolds number is between 10 and around 50, periodic velocity fluctuations are found in the jet. They die out as they travel downstream without developing into irregular fluctuations. When the Reynolds number exceeds about 50, periodic fluctuations develop into irregular, turbulent fluctuations. The frequency of the periodic fluctuation is roughly proportional to the square of the jet velocity.

The stability of the jet against an artificially imposed disturbance was also investigated. Sound was used as an artificial disturbance. The disturbance is either amplified or damped in the jet depending on its frequency. The conventional stability theory was modified by considering the streamwise increase of Reynolds number. The experimental results are in agreement with the theoretical results.

The stability of a plane liquid sheet is studied experimentally and theoretically, with an emphasis on the effect of the surrounding gas. Co-blowing with a gas velocity of the same order of magnitude as the liquid velocity is studied, in order to quantify its effect on the stability of the sheet. Experimental results are obtained for a water sheet in air at Reynolds number Rel = 3000 and Weber number We = 300, based on the half-thickness of the sheet at the inlet, water mean velocity at the inlet, the surface tension between water and air and water density and viscosity. The sheet is excited with different frequencies at the inlet and the growth of the waves in the streamwise direction is measured. The growth rate curves of the disturbances for all air flow velocities under study are found to be within 20% of the values obtained from a local spatial stability analysis, where water and air viscosities are taken into account, while previous results from literature assuming inviscid air overpredict the most unstable wavelength with a factor 3 and the growth rate with a factor 2. The effect of the air flow on the stability of the sheet is scrutinized numerically and it is concluded that the predicted disturbance growth scales with (i) the absolute velocity difference between water and air (inviscid effect) and (ii) the square root of the shear from air on the water surface (viscous effect).

Studies of the morphology of fragmentation of liquid drops in a liquid medium due to shock-induced flows are reported. For such high-density-ratio and high-interfacialtension systems, the fragmentation is found to be primarily due to the penetration of the drop by unstable waves. An envelope of conditions which characterize drop breakup in these systems has been deduced. Analytical models, which involve the time constants of both the drop-piercing Taylor instability and the wake-forming shear effects, reflect our experimental results well. In addition, it is found that the mercury/water system studied here should simulate the uranium dioxide/sodium pair, which is of interest in safety studies of fast reactors.