The instant of transition to cellularity of centrally ignited, outwardly propagating spherical flames in a reactive environment of fuelx–oxidizer mixture, at atmospheric and elevated pressures, was experimentally determined using high-speed schlieren imaging and subsequently interpreted on the basis of hydrodynamic and diffusional–thermal instabilities. Experimental results show that the transition Péclet number, Pec = RcℓL, assumes an almost constant value for the near-equidiffusive acetylene flames with wide ranges in the mixture stoichiometry, oxygen concentration and pressure, where Rc is the flame radius at transition and ℓL the laminar flame thickness. However, for the non-equidiffusive hydrogen and propane flames, Pec respectively increases and decreases somewhat linearly with the mixture equivalence ratio. Evaluation of Pec using previous theory shows complete qualitative agreement and satisfactory quantitative agreement, demonstrating the insensitivity of Pec to all system parameters for equidiffusive mixtures, and the dominance of the Markstein number, Ze(Le – 1), in destabilization for non-equidiffusive mixtures, where Ze is the Zel'dovich number and Le the Lewis number. The importance of using locally evaluated values of ℓL, Ze and Le, extracted from either computationally simulated one-dimensional flame structure with detailed chemistry and transport, or experimentally determined response of stretched flames, in the evaluation of Pec is emphasized.

Batchelor's vortex has been commonly used in the past as a model for aircraft trailing vortices. Using a temporal stability analysis, new viscous unstable modes have been found for the high swirl numbers of interest in actual large-aircraft vortices. We look here for these unstable viscous modes occurring at large swirl numbers (q > 1.5), and large Reynolds numbers (Re >103), using a spatial stability analysis, thus characterizing the frequencies at which these modes become convectively unstable for different values of q, Re, and for different intensities of the uniform axial flow. We consider both jet-like and wake-like Batchelor's vortices, and are able to analyse the stability for Re as high as 108. We also characterize the frequencies and the swirl numbers for the onset of absolute instabilities of these unstable viscous modes for large q.

The propagation of small-amplitude modes in an inviscid but sheared subsonic mean flow inside a duct is considered. For isentropic flow in a circular duct with zero swirl and constant mean flow density the pressure modes are described in terms of the eigenvalue problem for the Pridmore-Brown equation with Myers' locally reacting impedance boundary conditions.

The key purpose of the paper is to extend the results of the numerical study of the spectrum for the case of lined ducts with uniform mean flow in Rienstra (Wave Motion, vol. 37, 2003b, p. 119), in order to examine the effects of the shear and wall lining. In the present paper this far more difficult situation is dealt with analytically. The high-frequency short-wavelength asymptotic solution of the problem based on the WKB method is derived for the acoustic part of the spectrum. Owing to the stiffness of the governing equations, an accurate numerical study of the spectral properties of the problem for mean flows with strong shear proves to be a non-trivial task which deserves separate consideration.

The second objective of the paper is to gain theoretical insight into the properties of the hydrodynamic part of the spectrum. An analysis of hydrodynamic modes both in the short-wavelength limit and for the case of the narrow duct is presented. For simplicity, only the hard-wall flow configuration is considered.

Noise generation is investigated in subsonic isothermal round jets at Mach numbers M = 0.6 and M = 0.9, with Reynolds numbers ReD = 1700 and ReD ≥ 105, using causality methods on data provided by large-eddy simulations. The correlations between broadband sound pressure signals and broadband turbulence signals along the jet axis and the shear layer are calculated. The normalized correlations are found to be significant between the pressure emitted in the downstream direction and centreline flow quantities. They are much smaller in the cases involving flow quantities along the shear layer, and fall for large emission angles. The maximum correlations obtained between centreline turbulence and downstream sound pressure are observed just at the end of the potential core for time delays corresponding to the times of propagation evaluated along ray paths. They also appear to be lower as the Mach number is reduced, and to be enhanced as the Reynolds number is decreased. These correlation levels can reasonably be attributed to the noise source which is predominant at small emission angles. This source is therefore located on the jet centreline at the end of the potential core, in a flow region which is shown to be characterized by a dominant Strouhal number over a large axial distance, by a strong level of intermittency, and by a high convection velocity. This supports the contention that the downstream jet-noise component is connected to the periodic and intermittent intrusion of vortical structures into the jet core.

A comprehensive study on confined swirling flows in an operational gas-turbine injector was performed by means of large-eddy simulations. The formulation was based on the Favre-filtered conservation equations and a modified Smagorinsky treatment of subgrid-scale motions. The model was then numerically solved by means of a preconditioned density-based finite-volume approach. Calculated mean velocities and turbulence properties show good agreement with experimental data obtained from the laser-Doppler velocimetry measurements. Various aspects of the swirling flow development (such as the central recirculating flow, precessing vortex core and Kelvin–Helmholtz instability) were explored in detail. Both co- and counter-rotating configurations were considered, and the effects of swirl direction on flow characteristics were examined. The flow evolution inside the injector is dictated mainly by the air delivered through the primary swirler. The impact of the secondary swirler appears to be limited.

Results are presented from experimental investigations into the motion of a heavy ellipsoid in a horizontal rotating cylinder, which has been completely filled with highly viscous fluid. The motion can be conveniently classified using the ratio between the maximum radius of curvature of the ellipsoid κmax and the radius of the drum Rd. If κmax < Rd the ellipsoid adopts a fixed position adjacent to the rising wall for a given cylinder rotation rate. The dependence of this position on wall speed is, surprisingly, independent of the ellipsoid's length, and a Stokes flow model has been developed which predicts both this independence and the speed for the limiting case of an ellipsoid adjacent to a vertical wall. If κ max < Rd the ellipsoid must tilt in order to maintain the maximum surface area in close proximity to the wall. Once tilted, a component of the viscous drag acts to laterally translate the ellipsoid from end to end of the drum. The ellipsoid with κmax = Rd adopts a series of fixed positions for most drum rotational rates but, between two regions of fixed-point behaviour, it undergoes a transition to oscillatory motion.

Six sets of particle image velocimetry (PIV) data from the bottom boundary layer of the coastal ocean are examined. The data represent periods of high, moderate and weak mean flow relative to the amplitude of wave-induced motion, which correspond to high, moderate and low Reynolds numbers based on the Taylor microscale (Re). The two-dimensional PIV velocity distributions enable spatial filtering to calculate some of the subgrid-scale (SGS) stresses, from which we can estimate the SGS dissipation, and evaluate the performance of typically used SGS stress models. The previously reported mismatch between the SGS and viscous dissipation at moderate and low Reynolds numbers appears to be related to the sparsity of large vortical structures that dominate energy fluxes.

Conditional sampling of SGS stresses and dissipation based on wave phase using Hilbert transforms demonstrate persistent and repeatable direct effects of large-scale but weak straining by the waves on the SGS energy flux at small scales. The SGS energy flux is phase-dependent, peaking when the streamwise-wave-induced velocity is accelerating, and lower when this velocity is decelerating. Combined with strain rate generated by the mean flow, the streamwise wave strain causes negative energy flux (backscatter), whereas the vertical wave strain causes a positive flux. The phase-dependent variations and differences between horizontal and vertical contributions to the cascading process extend to strains that are substantially higher than the wave-induced motion. These trends may explain the measured difference between spatial energy spectra of streamwise velocity fluctuations and spectra of the wall-normal component, i.e. the formation of spectral bumps in the spectra of the streamwise component at the wavenumbers for the transition between inertial and dissipation scales.

All the model coefficients of typical SGS stress models measured here are phase dependent and show similar trends. Thus, the variations of measured SGS dissipation with phase are larger than those predicted by the model variables. In addition, the measured coefficients of the static Smagorinsky SGS stress model decrease with decreasing turbulence levels, and increase with filter size. The dynamic model provides higher correlation coefficients than the Smagorinsky model, but the substantial fluctuations in their values indicate that ensemble averaging is required. The ‘global’ dynamic model coefficients indicate that the use of a scale-dependent dynamic model may be appropriate. The structure function model yields poor correlation coefficients and is found to be over-dissipative under all but the highest turbulence levels. The nonlinear model has higher correlations with measured stresses, as expected, but it also does not reproduce the trends with wave phase.

Motivated by potential applications for offshore airports supported on vertical piles, we report a theory of wave diffraction by a periodic array of circular cylinders. The simple case of normal incidence on a rectangular array is studied here, which is equivalent to a line array along the centre of a long channel. An asymptotic theory is developed for cylinders much smaller than the incident wavelength, which is comparable to the cylinder spacing. Focus is on Bragg resonance near which scattering is strong. A combination of the method of multiple scales and the Bloch theorem leads to simple evolution equations coupling the wave envelopes. Dispersion of transient wave envelopes is investigated. Scattering of detuned waves by a large but finite number of cylinders is investigated for frequencies in and outside the band gap. Quantitative accuracy is assessed by comparisons with numerical computations via finite elements. The analytical theory prepares the ground for nonlinear studies and may facilitate future inclusion of real-fluid effects such as vortex shedding.

The statistics of a passive scalar randomly emitted from a point source is investigated analytically for the Kraichnan ensemble. Attention is focused on the two-point equal-time scalar correlation function, a statistical indicator widely used both in experiments and in numerical simulations. The only source of inhomogeneity/anisotropy is the injection mechanism, the advecting velocity being here statistically homogeneous and isotropic. The main question we address is on the possible existence of an inertial range of scales and a consequent scaling behaviour. The question arises from the observation that for a point source the injection scale is formally zero and the standard cascade mechanism cannot thus be taken for granted. We find from first principles that an intrinsic integral scale, whose value depends on the distance from the source, emerges as a result of sweeping effects. For separations smaller than this integral scale a standard forward cascade occurs. This is characterized by a Kolmogorov–Obukhov power-law behaviour as in the homogeneous case, except that the dissipation rate is also dependent on the distance from the source. Finally, we also find that the combined effect of a finite inertial-range extent and of inhomogeneities causes the emergence of subleading anisotropic corrections to the leading isotropic term, that are here quantified and discussed.

Detailed instantaneous velocity fields of a jet in crossflow have been measured with stereoscopic particle image velocimetry (PIV). The jet originated from a fully developed turbulent pipe flow and entered a crossflow with a turbulent boundary layer. The Reynolds number based on crossflow velocity and pipe diameter was 2400 and the jet to crossflow velocity ratios were R=3.3 and R=1.3. The experimental data have been analysed by proper orthogonal decomposition (POD). For R=3.3, the results in several different planes indicate that the wake vortices are the dominant dynamic flow structures and that they interact strongly with the jet core. The analysis identifies jet shear-layer vortices and finds that these vortical structures are more local and thus less dominant. For R=1.3, on the other hand, jet shear-layer vortices are the most dominant, while the wake vortices are much less important. For both cases, the analysis finds that the shear-layer vortices are not coupled to the dynamics of the wake vortices. Finally, the hanging vortices are identified and their contribution to the counter-rotating vortex pair (CVP) and interaction with the newly created wake vortices are described.

Detailed comparison of spatial direct numerical simulations (DNS) and secondary linear stability theory (SLST) is provided for the three-dimensional crossflow-dominated boundary layer also considered at the DLR-Göttingen for experiments and theory. Secondary instabilities of large-amplitude steady and unsteady crossflow vortices arising from one single primary mode have been analysed. SLST results have been found to be reliable with respect to the dispersion relation and the amplitude distribution of the modal eigenfunction in the crosscut plane. However, significant deviations have been found in the amplification rates, the SLST results being strongly dependent on the necessarily simplified representation of the primary state. The secondary instability mechanisms are shown to be local, i.e. robust with respect to violations of the periodicity assumption made in the SLST for the wall-parallel directions. Perturbations associated with different local maxima of the spanwise periodic eigenfunctions develop independently from each other interacting only with the primary vortices next to them. Characteristic structures induced by different secondary instability modes have been analysed and an analogy with the Kelvin–Helmholtz instability mechanism has been highlighted.

Droplet segregation in isotropic homogeneous turbulence is analysed using a spectral direct numerical simulation solver to describe the evolution of the turbulent carrier phase, whose characteristic properties remain statistically stationary due to a semi-deterministic forcing scheme. Lagrangian dilute spray modelling is employed to describe the discrete-phase evolution. The liquid density is distributed on an Eulerian mesh to analyse the evolution of the spray and its spatial distribution. This gives results in accordance with classical methods for droplet segregation. It also allows a deeper analysis of the spray evolution. In particular, droplet segregation and vapour mass fraction may be analysed jointly. First, droplet segregation phenomena are studied through the analysis of the formation and the geometry of the droplet clusters. Then, the effects of segregation on spray evaporation are investigated from both the dispersed and carrier phase points of view. At equilibrium, droplet dynamics leads to different segregation levels that are associated with characteristic Stokes numbers. It appears that the evaporation process evolves in three different stages in time: single-droplet mode in the early stage, cluster mode in the intermediate stage and a gaseous mode in the late stage. Segregation levels strongly affect the evolution of the mean vapour mixture fraction during the second stage, while the corresponding standard deviation is affected for longer, up to the third stage in our simulations. However, from the evolution of the integral scale and the shape of the energy spectrum, it appears that turbulent mixing eliminates the segregation effects, apart from the first evaporation stage when the droplet segregation determines the vapour distribution.

Bounds on the momentum transport by laminar or turbulent shear flows between two parallel plates in constant relative motion in a rotating system are derived. The axis of rotation is parallel to the plates. The dimensionless component of the rotation vector perpendicular to the relative motion of the plate is denoted by the Coriolis number τ. Through the consideration of separate energy balances for the poloidal and the toroidal components of the fluid velocity field a variational problem is formulated in which τ enters as a parameter. Bounds that are derived under the hypothesis that the extremalizing vector fields are independent of the streamwise coordinate suggest that no state of turbulent motion can exist for with .

In this paper, we use numerical simulation and laboratory experimental observation to show that fire whirls can be generated spontaneously through the interaction between a central flame and surrounding organized or randomly distributed flames. The momentum of the air stream entrained by the main flame decreases as it crosses a surrounding flame, so that the main flame rotates if surrounding flames are arranged in such a way as to block the passage of the air stream directed towards the centre of the main flame and to favour flows in a particular circumferential direction. An analysis is performed to study the role of the rotation speed in the flame height. It is found that the flame height initially decreases to a minimum owing to the inflow boundary layer wind reducing the initial vertical velocity of gas for low rotation speed and to entrainment enhancement reducing the rising time, and then it increases owing to the pressure reduction at the centre of the rotating vortex and entrainment suppression extending the rising time.

We study the electrohydrodynamic stability of the interface between two superposed viscous fluids in a channel subjected to a normal electric field. The two fluids can have different densities, viscosities, permittivities and conductivities. The interface allows surface charges, and there exists an electrical tangential shear stress at the interface owing to the finite conductivities of the two fluids. The long-wave linear stability analysis is performed within the generic Orr–Sommerfeld framework for both perfect and leaky dielectrics. In the framework of the long-wave linear stability analysis, the wave speed is expressed in terms of the ratio of viscosities, densities, permittivities and conductivities of the two fluids. For perfect dielectrics, the electric field always has a destabilizing effect, whereas for leaky dielectrics, the electric field can have either a destabilizing or a stabilizing effect depending on the ratios of permittivities and conductivities of the two fluids. In addition, the linear stability analysis for all wavenumbers is carried out numerically using the Chebyshev spectral method, and the various types of neutral stability curves (NSC) obtained are discussed.

The unfolding of inertial instability in intially barotropic vortices in a uniformly rotating and stratified fluid is studied through numerical simulations. The vortex dynamics during the instability is examined in detail. We demonstrate that the instability is stabilized via redistribution of angular momentum in a way that produces a new equilibrated barotropic vortex with a stable velocity profile. Based on extrapolations from the results of a series of simulations in which the Reynolds number and strength of stratification are varied, we arrive at a construction based on angular momentum mixing that predicts the infinite-Reynolds-number form of the equilibrated vortex toward which inertial instability drives an unstable vortex. The essential constraint is conservation of total absolute angular momentum. The construction can be used to predict the total energy loss during the equilibration process. It also shows that the equilibration process can result in anticyclones that are more susceptible to horizontal shear instabilities than they were initially, a phenomenon previously observed in laboratory and numerical studies.

Inertial instability in a rotating shear flow redistributes absolute linear momentum in such a way as to neutralize the instability. In the absence of other instabilities, the final equilibrium can be predicted by a simple construction based on conservation of total momentum. Numerical simulations, invariant in the along-stream direction, suppress barotropic instability and allow only inertial instability to develop. Such simulations, at high Reynolds numbers, are used to test the theoretical prediction. Four representative examples are given: a jet, a wall-bounded jet, a mixing layer and a wall-bounded shear layer.

The formation process of a non-classical rarefaction shock wave in dense gas shock tubes is investigated by means of numerical simulations. To this purpose, a novel numerical scheme for the solution of the Euler equations under non-ideal thermodynamics is presented, and applied for the first time to the simulation of non-classical fully three-dimensional flows. Numerical simulations are carried out to study the complex flow field resulting from the partial burst of the shock tube diaphragm, a situation that has been observed in preliminary trials of a dense gas shock tube experiment. Beyond the many similarities with the corresponding classical flow, the non-classical wave field is characterized by the occurrence of anomalous compression isentropic waves and rarefaction shocks propagating past the leading rarefaction shock front. Negative mass flow through the rarefaction shock wave results in a limited interaction with the contact surface close to the diaphragm, a peculiarity of the non-classical regime. The geometrical asymmetry does not prevent the formation of a single rarefaction shock front, though the pressure difference across the rarefaction wave is predicted to be weaker than the one which would be obtained by the complete burst of the diaphragm.

The structure of a columnar vortex in a stably stratified fluid is studied experimentally and theoretically when the vortex axis is slightly tilted with respect to the direction of stratification. When the Froude number of the vortex is larger than 1, we show that tilting induces strong density variations and an intense axial flow in a rim around the vortex. We demonstrate that these characteristics can be associated with a critical-point singularity of the correction of azimuthal wavenumber m = 1 generated by tilting where the angular velocity of the vortex equals the Brunt–Väisälä frequency of the stratified fluid. The theoretical structure obtained by smoothing this singularity using viscous effects (in a viscous critical-layer analysis) is compared to particle image velocimetry measurements of the axial velocity field and visualizations of the density field and a good agreement is demonstrated.

A striking resemblance between the normal-field instability in ferromagnetic fluids and the interfacial mode of the thermocapillary instability in viscous fluids is presented. A nonlinear evolution equation describing the dynamics of the free surface for a ferrofluid layer subject to a uniform normal magnetic field is derived, and compared to that for a thin viscous layer heated from below. Their similarity predicts the possibility of mutual nonlinear stability control.