Symmetry reduction by the method of slices is applied to pipe flow in order to obtain a quotient of the streamwise translation and azimuthal rotation symmetries of turbulent flow states. Within the symmetry-reduced state space, all travelling wave solutions reduce to equilibria, and all relative periodic orbits reduce to periodic orbits. Projections of these solutions and their unstable manifolds from their infinite-dimensional symmetry-reduced state space onto suitably chosen two- or three-dimensional subspaces reveal their interrelations and the role they play in organizing turbulence in wall-bounded shear flows. Visualizations of the flow within the slice and its linearization at equilibria enable us to trace out the unstable manifolds, determine close recurrences, identify connections between different travelling wave solutions and find, for the first time for pipe flows, relative periodic orbits that are embedded within the chaotic saddle, which capture turbulent dynamics at transitional Reynolds numbers.

In Marstorp et al. (J. Fluid Mech., vol. 639, 2009, pp. 403–432), an explicit algebraic subgrid stress model (EASSM) for large-eddy simulation (LES) was proposed, which was shown to considerably improve LES predictions of rotating and non-rotating turbulent channel flow. In this paper, we extend that work and present a new explicit algebraic subgrid scalar flux model (EASSFM) for LES, based on the modelled transport equation of the subgrid-scale (SGS) scalar flux. The new model is derived using the same kind of methodology that leads to the explicit algebraic scalar flux model of Wikström et al. (Phys. Fluids, vol. 12, 2000, pp. 688–702). The algebraic form is based on a weak equilibrium assumption and leads to a model that depends on the resolved strain-rate and rotation-rate tensors, the resolved scalar-gradient vector and, importantly, the SGS stress tensor. An accurate prediction of the SGS scalar flux is consequently strongly dependent on an accurate description of the SGS stresses. The new EASSFM is therefore primarily used in connection with the EASSM, since this model can accurately predict SGS stresses. The resulting SGS scalar flux is not necessarily aligned with the resolved scalar gradient, and the inherent dependence on the resolved rotation-rate tensor makes the model suitable for LES of rotating flow applications. The new EASSFM (together with the EASSM) is validated for the case of passive scalar transport in a fully developed turbulent channel flow with and without system rotation. LES results with the new model show good agreement with direct numerical simulation data for both cases. The new model predictions are also compared to those of the dynamic eddy diffusivity model (DEDM) and improvements are observed in the prediction of the resolved and SGS scalar quantities. In the non-rotating case, the model performance is studied at all relevant resolutions, showing that its predictions of the Nusselt number are much less dependent on the grid resolution and are more accurate. In channel flow with wall-normal rotation, where all the SGS stresses and fluxes are non-zero, the new model shows significant improvements over the DEDM predictions of the resolved and SGS quantities.

A model is developed for describing mixing of several species under high-pressure conditions. The model includes the Peng–Robinson equation of state, a full mass-diffusion matrix, a full thermal-diffusion-factor matrix necessary to incorporate the Soret and Dufour effects and both thermal conductivity and viscosity computed for the species mixture using mixing rules. Direct numerical simulations (DNSs) are conducted in a temporal mixing layer configuration. The initial mean flow is perturbed using an analytical perturbation which is consistent with the definition of vorticity and is divergence free. Simulations are performed for a set of five species relevant to hydrocarbon combustion and an ensemble of realizations is created to explore the effect of the initial Reynolds number and of the initial pressure. Each simulation reaches a transitional state having turbulent characteristics and most of the data analysis is performed on that state. A mathematical reformulation of the flux terms in the conservation equations allows the definition of effective species-specific Schmidt numbers $(\mathit{Sc})$ and of an effective Prandtl number $(\mathit{Pr})$ based on effective species-specific diffusivities and an effective thermal conductivity, respectively. Because these effective species-specific diffusivities and the effective thermal conductivity are not directly computable from the DNS solution, we develop models for both of these quantities that prove very accurate when compared with the DNS database. For two of the five species, values of the effective species-specific diffusivities are negative at some locations indicating that these species experience spinodal decomposition; we determine the necessary and sufficient condition for spinodal decomposition to occur. We also show that flows displaying spinodal decomposition have enhanced vortical characteristics and trace this aspect to the specific features of high-density-gradient magnitude regions formed in the flows. The largest values of the effective species-specific $\mathit{Sc}$ numbers can be well in excess of those known for gases but almost two orders of magnitude smaller than those of liquids at atmospheric pressure. The effective thermal conductivity also exhibits negative values at some locations and the effective $\mathit{Pr}$ displays values that can be as high as those of a liquid refrigerant. Examination of the equivalence ratio indicates that the stoichiometric region is thin and coincides with regions where the mixture effective species-specific Lewis number values are well in excess of unity. Very lean and very rich regions coexist in the vicinity of the stoichiometric region. Analysis of the dissipation indicates that it is dominated by mass diffusion, with viscous dissipation being the smallest among the three dissipation modes. The sum of the heat and species (i.e. scalar) dissipation is functionally modelled using the effective species-specific diffusivities and the effective thermal conductivity. Computations of the modelled sum employing the modelled effective species-specific diffusivities and the modelled effective thermal conductivity shows that it accurately replicates the exact equivalent dissipation.

This paper focuses on the characterization and the stochastic modelling of the fluid acceleration in turbulent channel flow. In the first part, the acceleration is studied by direct numerical simulation (DNS) at three Reynolds numbers (${\mathit{Re}}_{\ast } = {u}_{\ast } h/ \nu = 180$, 590 and 1000). It is observed that whatever the wall distance is, the norm of acceleration is log-normally distributed and that the variance of the norm is very close to its mean value. It is also observed that from the wall to the centreline of the channel, the orientation of acceleration relaxes statistically towards isotropy. On the basis of dimensional analysis, a universal scaling law for the acceleration norm is proposed. In the second part, in the framework of the norm/orientation decomposition, a stochastic model of the acceleration is introduced. The stochastic model for the norm is based on fragmentation process which evolves across the channel with the wall distance. Simultaneously the orientation is simulated by a random walk on the surface of a unit sphere. The process is generated in such a way that the mean components of the orientation vector are equal to zero, whereas with increasing wall distance, all directions become equally probable. In the third part, the models are assessed in the framework of large-eddy simulation with stochastic subgrid acceleration model (LES–SSAM), introduced recently by Sabel’nikov, Chtab-Desportes & Gorokhovski (Euro. Phys. J. B, vol. 80, 2011, p. 177–187), and designed to account for the intermittency at subgrid scales. Computations by LES–SSAM and its assessment using DNS data show that the prediction of important statistics to characterize the flow, such as the mean velocity, the energy spectra at small scales, the viscous and turbulent stresses, the distribution of the acceleration can be considerably improved in comparison with standard LES. In the last part of this paper, the advantage of LES–SSAM in accounting for the subgrid flow structure is demonstrated in simulation of particle-laden turbulent channel flows. Compared to standard LES, it is shown that for different Stokes numbers, the particle dynamics and the turbophoresis effect can be predicted significantly better when LES–SSAM is applied.

Flow through axisymmetric and eccentric sinuous stenoses is investigated numerically, for Reynolds numbers up to 400. The eccentricity consists of an offset of the stenosis throat. A range of stenosis eccentricity is tested; the wake flow is found to be highly sensitive to small eccentricities in the stenosis geometry, even with stenosis offsets of the order of the machining precision of experimental test-sections. Comparisons are made between the numerically simulated flow through stenoses with small eccentricities and results from the literature of non-axisymmetric flows through nominally axisymmetric geometries. The effect of distortion to the inlet Poiseuille velocity profile is also investigated and found to have a significantly less severe effect on the downstream wake flow than geometric eccentricity.

We investigate the clustering morphology of a swarm of freely rising deformable bubbles. A three-dimensional Voronoï analysis enables us to distinguish quantitatively between two typical preferential clustering configurations: a regular lattice arrangement and irregular clustering. The bubble data are obtained from direct numerical simulations using the front-tracking method. It is found that the bubble deformation, represented by the aspect ratio $\chi $, plays a significant role in determining which type of clustering is realized: nearly spherical bubbles form a regular lattice arrangement, while more deformed bubbles show irregular clustering. Remarkably, this criterion for the clustering morphology holds for different diameters of the bubbles, surface tensions and viscosities of the liquid in the studied parameter regime. The mechanism of this clustering behaviour is most likely connected to the amount of vorticity generated at the bubble surfaces.

We conduct a series of laboratory experiments to study the mixing of an initially linear stratification in turbulent Taylor–Couette flow. We vary the inner radius, ${R}_{1} $, and rotation rate, $\Omega $, relative to the fixed outer cylinder, of radius ${R}_{2} $, as well as the initial buoyancy frequency ${N}_{0} = \sqrt{(- g/ \rho )\partial \rho / \partial z} $. We find that a linear stratification spontaneously splits into a series of layers and interfaces. The characteristic height of these layers is proportional to ${U}_{H} / {N}_{0} $, where ${U}_{H} = \sqrt{{R}_{1} { \mathrm{\Delta} }_{R} } \Omega $ is a horizontal velocity scale, with ${ \mathrm{\Delta} }_{R} = {R}_{2} - {R}_{1} $ the gap width of the annulus. The buoyancy flux through these layers matches the equivalent flux through a two-layer stratification, independently of the height or number of layers. For a strongly stratified flow, the flux tends to an asymptotic constant value, even when multiple layers are present, consistent with Woods et al. (J. Fluid Mech., vol. 663, 2010, pp. 347–357). For smaller stratification the flux increases, reaching a maximum just before the layers disappear due to overturning of the interfaces.

The halting scale of the inverse energy cascade and the partition between kinetic and potential energy are considered for the case of forced quasigeostrophic turbulence in the regime of intermediate Rossby deformation length, for which the deformation length is comparable to the energy-containing scales of the flow. Phenomenological estimates for the halting scale and equilibrated energy of the forced–dissipative system with a simple representation of large-scale thermal damping are tested against numerical integrations and are found to poorly describe the numerically obtained dependence on damping coefficient; a modified scaling law is proposed that more accurately describes the dependence. The scale-selective nature of the damping leads to a large-scale spectral bottleneck that steepens the energy spectrum, consistent with previous studies of hypodiffusive dissipation. It is found that, across the parameter range considered, the blocking is largely insensitive to the ratio of deformation radius to the energy-containing scales.