The horizontally ($\mathrm{\Omega }\perp {\mathbf{v}}_{\mathrm{ns}}$) and axially ($\mathrm{\Omega }\parallel {\mathbf{v}}_{\mathrm{ns}}$) rotating counterflow of superfluid ${}^4\mathrm{He}$ (He II) generated thermally in a square channel is studied using the second sound attenuation technique, detecting the statistically steady state and temporal decay of the density of quantized vortex lines $L(t,\mathrm{\Omega })$. The array of rectilinear quantized vortices created by rotation at an angular velocity $\mathrm{\Omega }$ strongly affects the transient regimes of quantum turbulence characterized by the counterflow velocity ${\mathbf{v}}_{\mathrm{ns}}$, differently in both geometries. Two effects are observed, acting against each other and affecting the late temporal decay $L(t,\mathrm{\Omega })$. The first is the gradual decrease of the decay exponent $\mu$ of the power law $L(t,\mathrm{\Omega })\propto t^{-\mu }$, associated with the fact that under rotation the thermal counterflow acquires two-dimensional features, clearly observed and recently reported by us [Phys. Fluids 36, 105121 (2024)] in the $\mathrm{\Omega }\parallel {\mathbf{v}}_{\mathrm{ns}}$ geometry. It exists in the $\mathrm{\Omega }\perp {\mathbf{v}}_{\mathrm{ns}}$ geometry as well, however, it is screened here by the influence of the effective Ekman layer built within the effective Ekman time of order seconds. For faster rotation rates $L(t,\mathrm{\Omega })$ gradually ceases to display a clear power law. Instead, rounded and ever steeper decays occur, gradually shifting toward shorter and shorter times, significantly reducing the time range for a possible self-similar decay of the vortex line density. This effect is not observed in $\mathrm{\Omega }\parallel {\mathbf{v}}_{\mathrm{ns}}$ geometry, as here the much longer effective Ekman time of order minutes cannot affect the observed $L(t,\mathrm{\Omega })$ decay appreciably.

While flow optimization has been extensively studied in the continuum regime, its extension to rarefied gas flows remains less explored. Here, based on the Boltzmann model equation, an adjoint topology optimization method is employed to design two-dimensional single-inlet/multi-outlet manifolds, aiming to maximize the total mass flow rate while maintaining outflow uniformity. Two key findings are revealed. First, analogous to the Knudsen minimum in mass flow rate in the transition regime, a wetted-area minimum is identified, but in the slip flow regime. This phenomenon arises from the competition between flow bend loss and surface friction loss, with the latter being affected by velocity slip at the solid surface. Second, the inlet-outlet reciprocity emerges in the free-molecular flow regime, where the optimal design becomes invariant to inlet-outlet orientation and pressure ratio. Additional insights are gained regarding the channel curvature, compressibility effects, and the constraint of outflow uniformity. These findings elucidate the mechanisms governing rarefied gas transport and offer design guidance for manifolds operating in vacuum environments.

Modeling rarefied hypersonic flows remains a fundamental challenge due to the breakdown of classical continuum assumptions in the transition-continuum regime, where the Knudsen number ranges from approximately 0.1 to 10. Conventional Navier-Stokes-Fourier models with empirical slip-wall boundary conditions fail to accurately predict nonequilibrium effects such as velocity slip, temperature jump, and shock structure deviations. In this work, we develop a physics-constrained machine learning framework that augments bulk transport models and boundary conditions to extend the applicability of continuum solvers in nonequilibrium hypersonic regimes. We employ deep learning partial differential equation models for the viscous stress and heat flux embedded in the governing equations and trained via adjoint-based optimization. Three closure strategies are evaluated for two-dimensional supersonic flat-plate flows in argon across a range of Mach and Knudsen numbers. Additionally, we introduce a wall model based on a mixture of skewed Gaussian approximations of the particle velocity distribution function. This wall model replaces empirical slip conditions with physically informed, data-driven boundary conditions for the streamwise velocity and wall temperature. Our results show that a trace-free anisotropic viscosity model, paired with the skewed-Gaussian distribution function wall model, achieves significantly improved accuracy in predicting both bulk and wall quantities, particularly at high-Mach-number- and high-Knudsennumber regimes. Strategies such as parallel training across multiple Knudsen numbers and inclusion of high-Mach-number data during training are shown to enhance model generalization. Increasing model complexity yields diminishing returns for out-of-sample cases, underscoring the need to balance degrees of freedom and overfitting. To further assess generalization, we test the best-performing model on a geometrically out-of-sample wedge with increasing wall angles, for which the accuracy of the augmented predictions is consistent with the deviation from the training configuration. This work establishes data-driven, physics-consistent strategies for improving hypersonic flow modeling for regimes in which conventional continuum approaches are invalid.

Rayleigh-Bénard convection in a horizontal fluid layer is sustained only above a finite critical Rayleigh number. When the fluid is electrically conducting, a vertical magnetic field increases the critical Rayleigh number for the instability of the purely heat-conducting state in the bulk of the layer. When electrically insulating side walls are taken into account, the least stable mode becomes localized near these walls. Previous works studied the linear stability of such a wall-attached convection mode in circular containers or along a straight periodic wall. The present work considers a rectangular container, where induced eddy currents are presumably even more obstructed at the corners between insulating side walls. The linear stability problem is formulated using the quasistatic approximation of the induction equation. It is discretized and solved numerically for free-slip container walls by means of a two-dimensional spectral collocation method. The localization of the least stable mode at the corner is confirmed, and its properties are characterized by various geometric quantities. The effect of the magnetic field strength quantified by the Hartmann number is also examined, and similarities and differences between the wall-attached convection mode on a straight wall and in a corner are pointed out.

Wedge-shaped geometries in low-Reynolds-number flows are of increasing importance, for instance, in the design of microfluidic devices. The corresponding Green's functions describing the induced flow in response to a locally applied force were derived some time ago. To achieve a complete characterization of particle motion at low Reynolds numbers, we derive the flow response to locally applied torques. This is accomplished through a direct calculation based on the Fourier-Kontorovich-Lebedev transform using the Papkovich-Neuber representation of the hydrodynamic fields. We then illustrate the resulting flow fields, highlighting their structure, key features, and dependence on the geometry and orientation of the applied torque. Based on these solutions, we compute the corresponding hydrodynamic mobility tensor that couples torque and motion. Owing to the broken spatial symmetry imposed by the wedge-shaped confinement, a particle subjected to a torque will experience not only rotational motion, but also translational motion. These results provide analytical tools relevant for predicting and controlling particle behavior in confined microfluidic environments.

Order within a chaotic system is crucial for understanding, predicting, and controlling the dynamics and statistics of the system. In this study, we propose an objective identification method of coherent vortices in turbulence, where we identify the axis of tubular vortices without any adjustable parameters. Since the only necessary quantity for this method is the velocity gradient tensor, we can easily apply it to experimental data. To demonstrate the capability of the proposed method, we conduct wind tunnel experiments of a turbulent boundary layer, optimized to extract the largest vortices. By combining the proposed method with the scale decomposition of the velocity gradient tensor, we show that many of the three-dimensional structures of the largest vortices in the flow are hairpin shaped.

In stratified and rotating systems such as large scale geophysical flows, turbulent energy can cascade downscale, enhancing mixing and dissipation, or upscale, sustaining coherent structures such as eddies and gyres. Quantifying and, where possible, locally modulating the balance between these pathways could improve our ability to locally manage coastal mixing, pollution dispersion, and regional climate modeling. Here, we demonstrate in a 13-m-diameter rotating tank that the local direction and persistence of the energy cascade are governed by stress-strain alignment and can be locally modulated, based on a local Rossby number. A steady horizontal turbulent momentum jet injected at mid-depth was analyzed using high-resolution planar particle image velocimetry. After applying Gaussian filters over a range of scales, we computed subgrid-scale Reynolds stress and strain rate tensors and quantified their alignment angle. In nonrotating flows, the instantaneous energy flux is spatially intermittent but overall unbiased. As rotation increases, coherent inverse-cascade patches expand and their Lagrangian persistence grows. Statistical analyses show that faster rotation increases the fraction of inverse-cascade events, while making extreme bursts less frequent. These results support a tensor-geometry framework linking cascade direction to the angle between stress and strain eigenvectors and point toward practical strategies for engineering turbulent energy pathways in environmental and industrial applications.

The atomization of airway mucus during speech represents a primary mechanism for respiratory aerosol generation, which is central to airborne pathogen transmission. While extensive research has examined single-phase phonation aerodynamics, the multiphase dynamics of droplet formation has received less attention. This study aims to fill this critical gap by presenting a multiphase computational fluid dynamics (CFD) model of the human vocal folds. The dynamic formation and subsequent breakup of the mucus bridge are simulated using a volume-of-fluid method coupled with an analytical model describing the time-dependent shape of the glottal orifice. The investigation reveals that initial, small-scale stochastic perturbations on the mucus surface are essential for triggering a physically realistic, multimodal breakup, in contrast to the delayed, symmetric rupture observed in idealized smooth-film simulations. During vocal-folds opening, the bridge thickness exhibits exponential decay rupturing after $2.32\text{ms}$ at a dynamic aspect ratio of approximately 20, a value significantly exceeding the static stability limit. Furthermore, a parametric study reveals that increasing transglottal pressure from $491\text{Pa}$ to $1472\text{Pa}$, corresponding to louder phonation, accelerates the thinning process and leads to earlier initial rupture after $2.04\text{ms}$. This work provides fundamental insights into the precursors of respiratory aerosol generation

We study how the presence of a sweep angle affects the behavior of a shock-boundary layer interaction. The study is carried out using spanwise-periodic Large-eddy Simulations with a freestream Mach number of $M=2.7$ and a Reynolds number based on the momentum thickness of ${\mathrm{Re}}_{\theta }\simeq 3200$. Several sweep angle values are investigated, namely, $\beta =0^{\circ }, {15}^{\circ },{20}^{\circ }$, and ${40}^{\circ }$. The mean flow is first analyzed, showing an enlargement of the separation area, both in longitudinal and wall-normal directions when the sweep angle is increased. Strong flow deviation is also recovered downstream of the reattachment point for swept cases. With regard to the unsteady dynamics of the interaction, energy is found near the mean separation location for intermediate frequencies (corresponding to a Strouhal number ${\mathrm{St}}_{L_{\mathrm{sep}}}\approx 0.15$ for $\beta ={15}^{\circ }$ and ${20}^{\circ }$ and ${\mathrm{St}}_{L_{\mathrm{sep}}}\approx 0.23$ for $\beta ={40}^{\circ }$ in the wide domain). The latter are significantly higher than the very low frequencies ($0.01<{\mathrm{St}}_{L_{\mathrm{sep}}}<0.1$) usually associated with the unswept interaction. We relate these intermediate frequencies to spanwise-traveling structures, using the Proper Orthogonal Decomposition of velocity fluctuations in a volume located around the interaction region. The influence of the spanwise extent of the computational domain is also examined. This leads to the conclusion that a sufficiently wide computational domain is necessary to allow for the development of these structures and the appearance of the aforementioned intermediate frequencies.

In fresh water below the temperature of maximum density, the equation of state for density is a nonlinear function of temperature. Moreover, the maximum possible density differences are less than $0.1\mathrm{kg}{\mathrm{m}}^{-3}$. This implies that in the late winter and early spring most mid to high latitude lakes are in a regime of very weak buoyancy effects. We perform three-dimensional direct numerical simulations (DNS) to ascertain how the nonlinearity of the equation of state in such situations impacts the onset of shear instability and transition. We find that regardless of the smallness of the buoyancy term, accounting for buoyancy is essential in setting the timing and scale of three-dimensionalization. The nonlinearity of the equation of state manifests in setting the details of transition and mixing characteristics at moderate shear-layer based Reynolds number ($\text{Re}=60$), while for weak Reynolds number ($\text{Re}=45$) ignoring nonlinearity can change whether instability occurs at all. At Reynolds numbers greater than about 120, inertia dominates and the effects of both the nonlinear equation of state and the buoyancy term are minimized. We explore the effect of the commonly made approximation of setting the Prandtl Number ($\text{Pr}=\nu /\kappa$) to be 1 instead of the physical value of 12 in cold water. We observe that this common, though unnecessary, approximation also leads to a different onset and time evolution of the instability, and hence should be avoided in simulations of the cold water regime.

We present a three-dimensional (3D) numerical study of Rayleigh–Bénard convection (RBC) under a horizontal magnetic field in a small-aspect-ratio rectangular cell (2/3:8/3:1, $Pr=0.025$). The Rayleigh number ($Ra$) is fixed at $5\times {10}^5$ and ${10}^6$, while the Hartmann number ($Ha$) spans 3 orders of magnitude to provide a comprehensive view ($0\sim {10}^3$). With increasing magnetic field strength, the flow evolves from a 3D cellular structure to a quasi-two-dimensional (Q2D) rolls state. This transition reorients the main circulation plane and alters the dominant heat transport mechanism, leading to nonmonotonic variations in both Nusselt ($Nu$) and Reynolds ($Re$) numbers. The initial anomalous drop in $Nu$ under weak fields arises from disrupted vertical plume coherence. In contrast to large-aspect-ratio systems, where more rolls enhance heat transport, the small-aspect-ratio cell favors a single domain-spanning large-scale circulation (LSC) as the most efficient heat carrier. Flow reversals are observed at intermediate $Ha$, with both regular and irregular types. These reversals result from the growth of one vortex in a vertically stacked roll pair, differing from the classical reversal mechanism based on roll breakup and reconnection. Proper orthogonal decomposition (POD) analysis further confirms that mode interaction and competition play a crucial role in triggering the reversals. Based on the dissipation-dominance framework, we derive a transport scaling for the strong-field regime, $Nu\sim R{a}^{1/3}H{a}^{-1/6}, Re\sim R{a}^{2/3}H{a}^{-1/3}P{r}^{-1}$, and validate it against available RBC datasets, revealing a unified decay law of transport efficiency across different geometries. Furthermore, we extend elliptical instability theory to Q2D magnetoconvection, showing that stronger magnetic fields increase the stability threshold, allowing for more deformed elliptical rolls—findings consistent with observed flow topologies and dimensionality transitions.

We present a discussion of the enthalpy transport at the interface between two homothermal regions in a particle-laden turbulent flow, in order to quantify the role of inertial particles and of fluid-particle thermal interaction. We use a phase-space analysis of particles in a reduced phase space, and a formulation of a moment of total enthalpy integral (MTEI) applied to data from direct numerical simulations (DNS) at a Taylor microscale Reynolds number equal to 56 in a thermally coupled two-way regime. This allows us to identify the main features of particle statistics and to measure the effects of particle inertia and turbulent convection of the heat transfer, providing a quantitative and interpretable evaluation of the process which does not require any hypothesis of self-similarity. We show how particles dominate the transport mechanisms, bridging enthalpy across eddy structures, in particular when their Stokes number approaches unity, making them accumulate across temperature fronts.

This study examines the effect of mass ratio $\left(m^{*}\right)$ on vortex-induced vibrations of circular cylinders initially in tandem arrangement undergoing rigid collision. A finite volume approach with a diffuse interface immersed boundary method analyzes wake dynamics at an initial gap ratio $g/d=1.5$, over the reduced velocity range $3\le U^{*}\le 15$, and Reynolds number $\text{Re}=200$. The structural response reveals a monotonically increasing amplitude $\left(A_y^{*}\right)$ with $U^{*}$, indicative of lock-in within the synchronization regime before saturating beyond $U^{*}\approx 10$, while a systematic decrease in $A_y^{*}$ with rising $m^{*}$ highlights the greater inertial resistance of heavier structures. Results show a transition from chaotic wakes at $m^{*}=1$ to organized vortex patterns at higher $m^{*}$. At low $m^{*}$, collisions induce irregular shedding and nonlinear oscillations. As $m^{*}$ increases, wake patterns stabilize into periodic modes, leading to organized displacement. The upstream cylinder undergoes strong sign reversals in the time-average lift coefficient, while the downstream cylinder largely assumes the opposite sign due to gap-flow deflection and shielding, with matching signs occurring only at pinned, near side-by-side collision states where the wake momentarily reorganizes into a symmetric dipole. Across all $m^{*}$, the circulation $\left({\mathrm{\Gamma }}_z\right)$ strength increases with $U^{*}$ and peaks near $U^{*}\approx 6$, then declines, with heavier cylinders consistently yielding higher ${\mathrm{\Gamma }}_z$. Phase-space analysis shows irregular interactions at low $m^{*}$, stabilizing into consistent antiphase synchrony as $m^{*}$ increases.

The initial value problem is solved for the excitation of long surface gravity waves in a continuously sheared flow. This reveals the presence of a continuous spectrum alongside the standard normal modes of gravity wave propagation. An analytical similarity solution for the evolution of the free surface displacement from the continuous spectrum is found for the impulse response to surface excitation. It is demonstrated that the continuous spectrum contribution can be a significant fraction of the surface response, with the amplitude of the continuous spectrum exceeding that of the upstream gravity wave mode for Froude numbers of order unity. Thus, the continuous spectrum is shown to be a physically important part of the gravity wave response to forcing in sheared flows. The Landau damped mode description of the continuous spectrum is found to provide a link between methods using dispersion relations for phase speeds within the range of the velocity profile and the variable-shear profiles that do not admit normal modes in this range.

Shearless turbulent/nonturbulent mixing has primarily been explored in the incompressible limit for nominally planar interfaces. In this work, we investigate compressible and geometrical effects on turbulent/nonturbulent mixing in the absence of a mean shear. Focusing on initially homogeneous isotropic turbulence adjacent to a quiescent fluid, we theoretically predict the evolution of the mixing region width and turbulent kinetic energy and validate these predictions using direct numerical simulation. For the planar case, we find that the mixing region grows as a power law in time ($\sim t^{2/3}$), consistent with incompressible theory. However, we demonstrate that the turbulent kinetic energy is affected by energy transport away from the mixing region due to dilatation, in addition to viscous dissipation. We establish the dependence of the time evolution of the turbulent kinetic energy on the parameters governing the problem (Taylor scale, rms velocity, and initial dilatation). Dilatation leads to an increased decay rate of turbulent kinetic energy. When considering a cylindrical geometry, the mixing region growth is self-similar with the mixing region width exhibiting a power-law dependence in time as $t^{1/2}$. The scaling for the turbulent kinetic energy decay can also be extended to include energy transport away from the mixing region by dilatation and taking into account the diverging geometry.

The manipulation of micro- and nanoparticles at fluid interfaces is central to optofluidics, microfluidics, and colloidal assembly. While capillarity-driven migration and ordering are well established, achieving stable and reconfigurable particle trapping remains a challenge. Here, we demonstrate that a neck-shaped meniscus formed around a vertically actuated rod acts as a dynamically tunable capillary well, enabling reversible particle confinement, directed migration, and release. Furthermore, elliptical rods create anisotropic meniscus curvature that guides particles toward regions of higher curvature; we identify an optimal aspect ratio of the rod cross-section that maximizes trapping efficiency. Experiments supported by simulations confirm the predicted critical meniscus height and quasistationary regime velocity. These findings establish a geometry-controlled, field-free strategy for robust and reconfigurable particle manipulation at fluid interfaces, extending the principles of curvature-guided assembly.

Clogging can occur in confined flows of particulate suspensions that carry either too many particles or particles that are too large or too sticky. As a result, clogging is a significant problem in many engineering systems, ranging from inkjet printer nozzles to the progressive clogging of irrigation systems, which impairs their performance and sometimes leads to the entire failure of a system. The challenge in studying clogging by suspensions of particles is that the underlying physics is complex and spans many length scales and timescales, such that developing a general framework remains complex. In this article, we discuss how and why flowing particles can get stuck and clog fluid systems. In particular, we highlight the role of various clogging mechanisms at play in different systems and recent efforts to characterize, model, and prevent, or at least delay, clogging. Predicting when clogging is likely to occur and working to prevent it can lead to the development of new design principles for clog-resilient systems, thereby improving the reliability of fluid systems.

This study demonstrates that the free-surface flow dynamics of dense piles of contactless silica microparticles depend on the resting period prior to flow. Microfluidic rotating drum experiments reveal that longer resting times lead to delayed flow onsets and reduced flow velocities, both evolving logarithmically with the resting time. These aging effects are more pronounced for thermally driven creep flows in piles with initial tilting angle below the athermal angle of repose, in contrast to piles initially tilted above this repose angle, where gravity-driven flows tend to gradually erase aging effects. Moreover, we show that the packing fraction does not change during the resting period, and that aging occurs in both monodisperse and polydisperse piles, indicating that crystallization is not required for the time-dependent behavior to appear. Remarkably, vigorous agitation that redisperses the particles fully restores the piles to their initial state, demonstrating that the observed effects are not due to sample degradation. These findings evidence a form of aging in quiescent suspensions intermediate between colloidal and granular media, where thermal fluctuations, still significant relative to particle weight, progressively stabilize the system, making it more resistant to flow and deformation.

The aerobreakup dynamics of parallel droplets subjected to shock-induced airflow is investigated through experiments, in which the parallel-arranged droplets with controllable size and spacing are generated through forced breakup of a free liquid jet. Utilizing high-speed photography, the breakup process of parallel droplets over wide ranges of spacing ratio and Weber number can be captured. At low Weber numbers, the droplet fragmentation mode shifts gradually from the traditional bag breakup to the trailing and shuttlecock modes as the spacing ratio decreases. At relatively high Weber numbers, the breakup mode changes from an open to a closed configuration, depending on whether the channel between adjacent droplets closes during the deformation process. A phase diagram of droplet fragmentation modes is constructed in the parametric region of Weber number versus spacing ratio, with the boundaries between different modes clarified theoretically. Quantitative analyses on droplet deformation further demonstrate that reducing the spacing ratio could accelerate filament formation and facilitate channel closure. Overall, this work offers insights into the interaction dynamics of droplet clusters, thereby providing useful guidance for applications such as fuel atomization.

A study of the stationary flow in a channel of flattened triangular cross section subject to evaporation on one end and in contact with a liquid reservoir on the other end is presented taking into account the liquid film flow in the corners of the channel and in the adsorbed thin liquid film together with transport by diffusion of the vapor in the gas phase. Three main regimes are distinguished depending on the inlet liquid pressure, namely, the pure corner flow regime, the regime with partial bulk invasion, and the regime where the channel is fully invaded by the liquid. The various regimes are summarized in a phase diagram that depend on the inlet liquid pressure and external air relative humidity. Models are derived for each regime combining the physical description of each transport mechanism. Numerical solutions are obtained using a simple bisection method. Results show that the mass transfer rate through the channel can vary over several orders of magnitude depending on the inlet pressure. Situations where the mass transfer rate is independent of the external air relative humidity are exhibited both for the pure corner flow and partial bulk invasion regimes.