Premixed turbulent flames, encountered in power generation and propulsion engines, are an archetype of a randomly advected, self-propagating surface. While such a flame is known to exhibit large-scale intermittent flapping, the possible intermittency of its small-scale fluctuations has been largely disregarded. Here, we experimentally reveal the inner intermittency of a premixed turbulent V-flame, while clearly distinguishing this small-scale feature from large-scale outer intermittency. From temporal measurements of the fluctuations of the flame, we find a frequency spectrum that has a power-law subrange with an exponent close to $-2$, which is shown to follow from Kolmogorov phenomenology. Crucially, however, the moments of the temporal increment of the flame position are found to scale anomalously, with exponents that saturate at higher orders. This signature of small-scale inner intermittency is shown to originate from high-curvature, cusp-like structures on the flame surface, which have significance for modelling the heat release rate and other key properties of premixed turbulent flames.

Experiments are reported that explore the onset of motion of bubbles in a model yield stress fluid, Carbopol gel. Starting from a trapped spherical bubble in a gel, the yielding limit for the bubble motion is obtained by gradually expanding the bubble via a stepwise decrease in pressure. Our results show that at the yielding limit bubbles are longer and thinner when they are in a higher concentrated gel. This is suggestive of a link between the shape and size of the bubbles at the onset of motion and the rheology of the material, in particular elastic behaviour below the yielding point. Particular attention has been paid to investigating the dynamic response of gel during the bubble growth. Subjecting the bubble to a periodic change in the pressure confirms the irreversibility of the gel deformation and its hysteresis, which are hallmarks of nonlinear viscoelastic behaviour of the gel before yielding. In this context, the periodic expansion and contraction of the bubbles leave residual deformation (stresses) in the gel which facilitates the liberation of bubbles.

We examine the effect of Dean number on the inertial focusing of spherical particles suspended in flow through curved microfluidic ducts. Previous modelling of particle migration in curved ducts assumed the flow rate was small enough that a leading-order approximation of the background flow with respect to the Dean number produces a reasonable model. Herein, we extend our model to situations involving a moderate Dean number (in the microfluidics context) while the particle Reynolds number remains small. Variations in the Dean number cause a change in the axial velocity profile of the background flow which influences the inertial lift force on a particle. Simultaneously, changes in the cross-sectional velocity components of the background flow directly affect the secondary flow induced drag. In keeping the particle Reynolds number small, we continue to approximate the inertial lift force using a regular perturbation while capturing the subtle effects from the modified background flow. This approach pushes the limits at which a regular perturbation is applicable to provide some insights into how variations in the Dean number influence particle focusing. Our results illustrate that, as the extrema in the background flow move towards the outside of edge of the cross-section with increasing Dean number, we observe a similar shift in the stable equilibria of some, but not all, particle sizes. This might be exploited to enhance the lateral separation of particles by size in a number of practical scenarios.

Responses of a geometry-induced separation bubble (GISB) and a pressure-induced separation bubble (PISB) at enhanced levels of free-stream turbulence (FST) have experimentally been investigated for a comparative study using the particle image velocimetry (PIV) technique. The outlines of separation bubbles based on the dividing streamlines are self-similar for different levels of FST and Reynolds numbers. The spectral analyses of the time-resolved PIV data show that the vortex shedding frequency of a separated shear layer remains unchanged for the GISB cases even with an enhanced level of FST. In contrast, it is different for the PISB cases. We propose a criterion that determines whether the frequency will remain the same even for the cases with FST. Linear stability analyses reveal that the inviscid-inflectional instability dominates the transition process, and the linear stages of transition are not completely bypassed even at an enhanced level of FST. The most amplified frequencies, while scaling with the displacement thickness and the boundary layer edge velocity, collapse in a single curve for all the cases. Furthermore, measurements in the spanwise plane show that the streamwise velocity streak/Klebanof mode at an enhanced level of FST is not a general flow feature for all types of separation bubbles. However, at an enhanced level of FST for the PISB case, the boundary layer streaks are found to distort the two-dimensional vortex structure associated with the Kelvin–Helmholtz instability, eventually leading to a three-dimensional $\varLambda$-like structure in the spanwise plane.

A closed-form solution is presented for internal compressible flows with an arbitrary combination of area change, heat addition, friction and non-uniform flow. The approach assumes steady flow of a calorically perfect ideal gas using the integral form of the governing equations. It is shown that the equations reduce to a simple, bi-quadratic equation in exit Mach number. This solution is compared to previously known solutions for Rayleigh flow, Fanno flow, isentropic flow and normal shock relations. Theoretical solutions for sudden expansion, sudden contraction, and supersonic–supersonic two-stream constant area mixing are also presented. These solutions were found to be in agreement with data available in the literature.

We experimentally explored the effect of single-sidewall cooling on Rayleigh–Bénard (RB) convection. Canonical RB was also studied to aid insight. The scenarios shared tank dimensions and bottom and top wall temperatures; the single sidewall cooling had the top wall temperature. Turbulence was explored at two canonical Rayleigh numbers, $Ra=1.6\times 10^{10}$ and $Ra=2\times 10^9$ under Prandtl number $Pr=5.4$. Particle image velocimetry described vertical planes parallel and perpendicular to the sidewall cooling. The two $Ra$ scenarios reveal pronounced changes in the flow structure and large-scale circulation (LSC) due to the sidewall cooling. The density gradient induced by the sidewall cooling led to asymmetric descending and ascending flows and irregular LSC. Flow statistics departed from the canonical case, exhibiting lower buoyancy effects, represented by an effective Rayleigh number with effective height dependent on the distance from the lateral cooling. Velocity spectra show two scalings, $\varPhi \propto f^{-5/3}$ Kolmogorov (KO41) and $\varPhi \propto f^{-11/5}$ Bolgiano (BO59) in the larger $Ra$; the latter was not present in the smaller set-up. The BO59 scaling with sidewall cooling appears at higher frequencies than its canonical counterpart, suggesting weaker buoyancy effects. The LSC core motions allowed us to identify a characteristic time scale of the order of vortex turnover time associated with distinct vortex modes. The velocity spectra of the vortex core oscillation along its principal axis showed a scaling of $\varPhi _c \propto f^{-5/3}$ for the single sidewall cooling, which was dominant closer there. It did not occur in the canonical case, evidencing the modulation of LSC oscillation on the flow.

Predicting the performance of large turbine arrays requires the understanding of many physical factors, such as array geometry, turbine operation, inflow conditions and turbulent wake mixing. Due to the large parameter space that an array may be optimised over, low-order models with low computational cost are often employed. This paper extends one of these models, the inviscid–viscous coupled model, for multi-row turbine modelling. Firstly, an extension to the inviscid actuator disc theory is presented by removing the limit on the number of discrete streamtubes computed. The extended model allows for the quantification of the impact of shear in the bypass and core flows separately. In particular, it is shown that averaging a sheared bypass flow profile can result in a substantial over-prediction of the power of a turbine in a laterally bounded flow as the effective blockage of the flow increases. The model is also used to confirm that an approximation using a limited number of streamtubes in some previous applications of the inviscid–viscous approach has a negligible impact on the results. Secondly, we explore the performance of a multi-row array with either uniform or varying turbine resistance across different rows. Results suggest that by varying resistance across rows, the array may outperform the uniform resistance case. The performance gain is dependent, however, on the arrangement and inter-turbine spacing both in the spanwise and streamwise directions.

When a viscous fluid spreads underneath a deformable surface skin or crust, the peeling dynamics at the fluid front can control the rate of advance rather than bulk self-similar flow. For an elastic skin, this control results in a quasi-static interior blister held at constant pressure that is matched to a narrow peeling region behind the fluid front. In this paper, the analogous problem is considered for a skin that deforms either viscously or plastically, or both. In particular, the deformable surface is assumed to be a thin plate of material governed by the Herschel–Bulkley constitutive law. We examine how such a skin controls viscous flow underneath, fed at constant flux and spreading as either a planar or axisymmetric current. As for an elastic skin, the peeling dynamics at the viscous fluid front again controls the rate of spreading. However, contrary to that situation, the mathematical matching problem for viscoplastic peeling is simplified considerably as a result of an integral constraint. Despite this, the structure of the peeling region is complicated significantly by any plasticity in the skin, which can create a convoluted peeling wave ahead of the main blister that features interwoven yielded and plugged sections of the plate.

We propose an analytical model to estimate the interface temperature $\varTheta _{\varGamma }$ and the Nusselt number $Nu$ for an evaporating two-layer Rayleigh–Bénard configuration in statistically stationary conditions. The model is based on three assumptions: (i) the Oberbeck–Boussinesq approximation can be applied to the liquid phase, while the gas thermophysical properties are generic functions of thermodynamic pressure, local temperature and vapour composition, (ii) the Grossmann–Lohse theory for thermal convection can be applied to the liquid and gas layers separately and (iii) the vapour content in the gas can be taken as the mean value at the gas–liquid interface. We validate this setting using direct numerical simulations in a parameter space composed of the Rayleigh number ($10^6\leq Ra\leq 10^8$) and the temperature differential ($0.05\leq \varepsilon \leq 0.20$), which modulates the variation of state variables in the gas layer. To better disentangle the variable property effects on $\varTheta _\varGamma$ and $Nu$, simulations are performed in two conditions. First, we consider the case of uniform gas properties except for the gas density and gas–liquid diffusion coefficient. Second, we include the variation of specific heat capacity, dynamic viscosity and thermal conductivity using realistic equations of state. Irrespective of the employed setting, the proposed model agrees very well with the numerical simulations over the entire range of $Ra$–$\varepsilon$ investigated.

We consider transverse shear flow over a $2L$-periodic array of air-filled grooves, which are separated by solid slats. We address the scenario where the liquid partially invades the grooves. Following Crowdy's solution of the longitudinal problem (J. Fluid Mech. vol. 925, 2021, R2), we focus upon the simplified geometry of infinitely narrow slats. For flat menisci, the ratio $\lambda$ of the slip length to $L$ depends upon a single geometric parameter, namely the ratio $H$ of the invasion depth to $L$. We analyse the singular limit $H\ll 1$ using matched asymptotic expansions, with an outer region on the scale of the period and an inner region on the scale of a single slat. In the outer region, we employ the complex-variable formulation of Stokes flows, obtaining the complex velocity as a divergent series. This formal representation is transformed to an admissible closed-form expression. In the inner region, the problem is equivalent to that of Stokes flow past a finite line segment. Asymptotic matching yields $\lambda = [\ln (2/{\rm \pi} H)-1]/{\rm \pi}$.

In this work, the stability of hypersonic flow over a curved compression ramp is studied using several stability analysis tools and direct numerical simulations (DNS). The free-stream Mach number and the unit Reynolds number are 7.7 and $4.2 \times 10^6$ m$^{-1}$, respectively. Corner rounding is considered to alter the separation bubble flow so as to suppress the intrinsic instability of the compression-ramp flow. The variation of intrinsic instability is confirmed by global stability analysis. Subsequently, resolvent analysis is employed to examine the response of intrinsically stable flows to external disturbances. It is shown that the considered flows strongly amplify low-frequency streamwise streaks with a preferential spanwise wavelength. This result is verified using DNS by introducing a random forcing upstream of the separation point. Furthermore, both resolvent analysis and DNS demonstrate that the separation bubble contributes little to the selection of the spanwise wavelength of streamwise streaks. The combined effects of convective and intrinsic instabilities are also explored using DNS. A better agreement with experimental data is achieved after introducing upstream disturbances in an inherently unstable flow.

This study proposes a newly developed deep-learning-based method to generate turbulent inflow conditions for spatially developing turbulent boundary layer (TBL) simulations. A combination of a transformer and a multiscale-enhanced super-resolution generative adversarial network is utilised to predict velocity fields of a spatially developing TBL at various planes normal to the streamwise direction. Datasets of direct numerical simulation (DNS) of flat plate flow spanning a momentum thickness-based Reynolds number, $Re_\theta = 661.5\unicode{x2013}1502.0$, are used to train and test the model. The model shows a remarkable ability to predict the instantaneous velocity fields with detailed fluctuations and reproduce the turbulence statistics as well as spatial and temporal spectra with commendable accuracy as compared with the DNS results. The proposed model also exhibits a reasonable accuracy for predicting velocity fields at Reynolds numbers that are not used in the training process. With the aid of transfer learning, the computational cost of the proposed model is considered to be effectively low. Furthermore, applying the generated turbulent inflow conditions to an inflow–outflow simulation reveals a negligible development distance for the TBL to reach the target statistics. The results demonstrate for the first time that transformer-based models can be efficient in predicting the dynamics of turbulent flows. They also show that combining these models with generative adversarial networks-based models can be useful in tackling various turbulence-related problems, including the development of efficient synthetic-turbulent inflow generators.

Diffusiophoresis takes place when a particle in solution moves due to the presence of a solute concentration gradient. This phenomenon is often studied under some simplifying assumptions, such as negligible diffusive layer thickness or infinite diffusion coefficient. In this work we simulate diffusiophoresis without these simplifications. The goal of this numerical study is to investigate equilibrium and fully developed states of non-electrolyte phoretic systems. Simulation results show that equilibrium states depend on solute diffusivity and on a reference solute concentration far from the particle. An expression is regressed that gives the (equilibrium) diffusiophoretic velocity as a function of solute concentration gradient, solute diffusion coefficient and the reference solute concentration far from the particle. A different set of results reveals that the state of phoretic systems does not depend on the initial conditions when time goes to infinity. This motivates the definition of fully developed states, designating those systems whose properties no longer depend on initial conditions. Apart from these findings, this work also depicts the effect of solute–interface interactions on diffusiophoresis. Simulation results for two solid particles with different interaction potentials are used to illustrate particle separation via diffusiophoresis. Finally, values of particle mobility are calculated for different solute–interface attraction strengths. These results are compared with another work in the literature, which studies polymer diffusiophoresis via molecular simulations (Ramírez-Hinestrosa et al., J. Chem. Phys., vol. 152, 2020, p. 164901).

The high-Reynolds-number stratified wake of a slender body is studied using a high-resolution hybrid simulation. The wake generator is a 6 : 1 prolate spheroid with a tripped boundary layer, the diameter-based body Reynolds number is ${Re}= U_\infty D/\nu = 10^5$, and the body Froude numbers are ${Fr}=U_\infty /ND=\{2,10,\infty \}$. The wake defect velocity decays following three stages with different wake decay rates (Spedding, J. Fluid Mech., vol. 337, 1997, pp. 283–301) as for a bluff body. However, the transition points among stages do not follow the expected $Nt = Nx/U_\infty$ values. Comparison with the wake of a circular disk in similar conditions (Chongsiripinyo & Sarkar, J. Fluid Mech., vol. 885, 2020) quantifies the influence of the wake generator – bluff versus slender – in stratified flow. The strongly stratified ${Fr}=2$ wake is in a resonant state. The steady lee waves strongly modulate the mean flow, and relative to the disk, the 6 : 1 spheroid (a high-aspect-ratio shape) wake at ${Fr}=2$ shows an earlier transition from the non-equilibrium (NEQ) stage to the quasi-two-dimensional (Q2D) stage. The NEQ–Q2D transition is followed by a sharp increase in the turbulent kinetic energy and horizontal wake meanders. At ${Fr}=10$, the start of the NEQ stage is delayed for the spheroid. Transfers between kinetic energy and potential energy reservoirs (both mean and turbulence) are analysed, and the flows are compared in phase space (with local Froude and Reynolds numbers as coordinates). Overall, the results of this study point to the difficulty of finding a universal framework for stratified wake evolution, independent of the features of the body, and provide insights into how buoyancy effects depend on the wake generator.

The stability and sensitivity of two- and three-dimensional global modes developing on steady spanwise-homogeneous laminar separated flows around NACA 4412 swept wings are numerically investigated for different Reynolds numbers ${\textit {Re}}$ and angles of attack $\alpha$. The wake dynamics is driven by the two-dimensional von Kármán mode whose emergence threshold in the $\alpha \unicode{x2013}{\textit {Re}}$ plane is computed with that of the three-dimensional centrifugal mode. At the critical Reynolds number, the Strouhal number, the streamwise wavenumber of the von Kármán mode and the spanwise wavenumber of the leading three-dimensional centrifugal mode scale as a power law of $\alpha$. The introduction of a sweep angle attenuates the growth of all unstable modes and entails a Doppler effect in the leading modes’ dynamics and a shift towards non-zero frequencies of the three-dimensional centrifugal modes. These are found to be non-dispersive as opposed to the von Kármán modes. The sensitivity of the leading global modes is investigated in the vicinity of the critical conditions through adjoint-based methods. The growth-rate sensitivity map displays a region on the suction side of the wing, wherein a streamwise-oriented force has a net stabilising effect, comparable to what could have been obtained inside the recirculation bubble. In agreement with the predictions of the sensitivity analysis, a spanwise-homogeneous force suppresses the Hopf bifurcation and stabilises the entire branch of von Kármán modes. In the limit of small amplitudes, passive control via spanwise-wavy forcing produces a stabilising effect similar to that of a spanwise-homogeneous control and is more effective than localised spherical forces.

We consider the Nusselt–Rayleigh number problem of Rayleigh–Bénard convection and make the hypothesis that the velocity and thermal boundary layer widths, $\delta _u$ and $\delta _T$, in the absence of a strong mean flow are controlled by the dissipation scales of the turbulence outside the boundary layers and, therefore, are of the order of the Kolmogorov and Batchelor scales, respectively. Under this assumption, we derive $Nu \sim Ra^{1/3}$ in the high $Ra$ limit, independent of the Prandtl number, $\delta _T/L \sim Ra^{-1/3}$ and $\delta _u/L \sim Ra^{-1/3} Pr^{1/2}$, where $L$ is the height of the convection cell. The scaling relations are valid as long as the Prandtl number is not too far from unity. For $Pr \sim 1$, we make a more general ansatz, $\delta _u \sim \nu ^{\alpha }$, where $\nu$ is the kinematic viscosity and assume that the dissipation scales as $\sim u^3/L$, where $u$ is a characteristic turbulent velocity. Under these assumptions we show that $Nu \sim Ra^{\alpha /(3-\alpha )}$, implying that $Nu \sim Ra^{1/5}$ if $\delta _u$ were scaling as in a Blasius boundary layer and $Nu \sim Ra^{1/2}$ (with some logarithmic correction) if it were scaling as in a standard turbulent shear boundary layer. It is argued that the boundary layers will retain the intermediate scaling $\alpha = 3/4$ in the limit of high $Ra$.

In particle-laden turbulent wall flows, transport of particles towards solid walls is phenomenologically thought to be governed by the wall-normal turbulence intensity supporting the underlying particle–eddy interactions that are usually modelled by a combination of turbophoresis and turbulent diffusion. We estimate the turbophoretic and turbulent diffusive coefficients as a function of wall-normal coordinate directly from a generated direct numerical simulation (DNS) database of low volume fraction point particles in a turbulent pipe flow. These coefficients are then used in an advection–diffusion equation to estimate the particle concentration as a function of wall-normal distance and time, with favourable comparison against DNS for smaller Stokes number ($St^+$) particles suggesting a limitation of the common gradient diffusion hypothesis for larger $St^+$ particles. Using DNS we explore the non-trivial effects of $St^+$, pipe wall condition (particle absorbing or elastic) as well as the influence of drag and lift force on the velocity and particle statistics giving rise to different particle concentrations. We then appraise various Eulerian-based models of turbophoretic and turbulent diffusive coefficients and, finally, use physical insights from Lagrangian correlation times, conditional quadrant analysis and flow topology to shed further light on the particle transport as a function of various parameters and the limits of gradient diffusion hypothesis.

The dynamics of sedimenting particles under gravity are surprisingly complex due to the presence of effective long-ranged forces. When the particles are polar with a well-defined symmetry axis and non-uniform density, recent theoretical predictions suggest that prolate objects will repel and oblate ones will weakly attract. We tested these predictions using mass polar prolate spheroids, which are composed of 2 mm spheres glued together. We probed different aspect ratios ($\kappa$) and centre of mass variations ($\chi$) by combining spheres of different densities. Experiments were done in both quasi-two-dimensional (2-D) and three-dimensional (3-D) chambers. By optically tracking the motion of single particles, we found that the dynamics were well described by a reduced mobility matrix model that could be solved analytically. Pairs of particles exhibited an effective repulsion, and their separation roughly scaled as $(\kappa - 1)/\chi ^{0.39}$, i.e. particles that were more prolate or had smaller mass asymmetry had stronger repulsion effects. In three dimensions, particles with $\chi >0$ were distributed more uniformly than $\chi =0$ particles, and the degree of uniformity increased with $\kappa$, indicating that the effective 2-body repulsion manifests for a large number of particles.

In this study, we revisit the spectral transfer model for the turbulent intensity in passive scalar transport (under large-scale anisotropic forcing), and a subsequent modification to the scaling of scalar variance cascade is presented. From the modified spectral transfer model, we obtain a revised scalar transport model using a fractional-order Laplacian operator that facilitates the robust inclusion of the non-local effects originating from large-scale anisotropy transferred across the multitude of scales in the turbulent cascade. We provide an a priori estimate for the non-local model based on the scaling analysis of the scalar spectrum, and later examine our developed model through direct numerical simulation. We present a detailed analysis on the evolution of the scalar variance, high-order statistics of the scalar gradient and important two-point statistical metrics of the turbulent transport to make a comprehensive comparison between the non-local model and its standard version. Finally, we present an analysis that seamlessly reconciles the similarities between the developed model with the fractional-order subgrid-scale scalar flux model for large-eddy simulation (Akhavan-Safaei et al., J. Comput. Phys., vol. 446, 2021, 110571) when the filter scale approaches the dissipative scales of turbulent transport. In order to perform this task, we employ a Gaussian process regression model to predict the model coefficient for the fractional-order subgrid model.

The heterogeneity of permeability in a porous medium considerably alters the behaviour of density-driven flows from what is observed in a medium of homogeneous permeability, and significantly enhances the mixing between the dense and light fluids during the flow. In this work, we present results from laboratory experiments performed in heterogeneous media consisting of horizontal layers of different permeabilities, investigating their effects on gravity current flows. We find that the mixing in our heterogeneous experimental set-ups can be ${O}(2)$ greater than that in a homogeneous medium of similar depth-averaged properties. The enhanced mixing in this setting is primarily because of transverse gravity-driven fingers and produced blunt front, which is the direct result of the layered structure. This enhanced rate of mixing dictates the gravity current height and length, making the current lose its long and thin shape much faster than a comparable current in a homogeneous medium. We discuss the experimental observations in detail and present relevant physical interpretations. Based on the experimental measurements and dimensionless modelling, we also derive semi-empirical formulas for predicting the gravity current length, height and mixing in a heterogeneous medium. Results from this work can be used in predicting the scale of mixing between different density fluids during contaminant transport in the subsurface environment.