Experiments find different forms of droplets that freeze on impacting a cold surface, including conical tipped (ICT), spherical caped (SCP), single torus (STS) and double torus (DTS).

A random forest algorithm is trained to recognize different transitions to turbulence in wavy-wall tubes, and then can classify a flow within 1/100th of a second.

A small change of slip boundary condition within a lubrication region breaks the fore-aft symmetry, which leads to a significant lift force. The change of slippage arises naturally on surfaces where physical and/or chemical properties are not perfectly constant. The induced lift force may result in nontrivial trajectories of particles traveling near surfaces.

Natural and microfluidic environments display a variety of confining surfaces with structured topographies, which modify the surrounding flow fields and impact nearby particle motion. We study a non-Brownian spherical particle sedimenting nearby a random, rough surface by using an analytical theory and numerical simulations. Roughness of the wall induces fluctuations in the velocity of the fall, leading to a quadratic increase of the variance of the displacements at long times.

A cobalt chloride membrane grows in a bath of sodium silicate in stable, oscillatory or explosively unstable regimes, with experiments confirming a simple model.

The epistemic uncertainty in turbulent scalar flux modeling directly affects the accuracy of scalar transport simulations. Two classical phenomenological theories on the pressure scrambling effect in scalar flux dynamics—return-to-isotropy and isotropization-of-production—are revisited from the perspective of uncertainty quantification (UQ). A prototype scalar flux model UQ method is thus formulated, providing a physical ground for the future data-driven UQ practice.

The propagation of random water waves is experimentally investigated in a 150-m-long flume. The emergence of high-amplitude localized structures by nonlinear self-focusing, which correlates to some statistical properties of surface elevation, is reported. A systematic study reveals to what extent these structures are locally akin to Peregrine solitons, a famous prototype of rogue waves.

A linear stability theory is developed to examine the conditions under which self-sustained oscillations occur in elastic filaments attached to a sphere or a wall subject to uniform force applied parallel to the axis of the filament. We shown that N equi-spaced filaments attached to a wall can undergo N different oscillation modes. The mode of first oscillation onset as the magnitude of the tangential force is increased, however, always corresponds to the mode of all filaments oscillating together in-phase. The figure shows the critical load and emergent frequency at the onset of instability for filaments oscillating in phase.

The rheology of suspensions of solid spheres in viscoelastic background media is analytically calculated for dilute values of solid volume fraction. A mean-field cell model is used to express the influence of the solid particles in a suspending fluid, which obeys the second-order fluid model. A semi-analytical constitutive equation is created for the whole suspension, itself a second-order fluid with modified material parameters. The resulting rheology of the whole suspension is presented in simple shear, which shows predictions for the behavior of viscometric functions past explicitly dilute concentrations.

Experimental evidence is provided for the authors’ theoretical prediction that suspended disks reduce the onset speed for instability in Taylor-Couette flow.

The microscopic structure of a physical gel such as Carbopol flowing in a confined microchannel is strongly heterogeneous over space. Thus tackling the wall-slip behavior using the classical assumption of the presence of a thin layer of Newtonian solvent of constant width near the channel’s walls is unjustified. To address the wall-slip scaling behavior, we propose a novel method of directly relating macroscopic rheological measurements to the measured wall velocity gradients. The scaling behavior we obtain via this method differs substantially from the previously published results.

The effect of vibrational nonequilibrium of SF${}_6$ molecules on the gas dynamic structure of microjets is revealed, manifesting in a decrease in the longitudinal cell size of the wave structure and weakening of variations in flow parameters along the axis of the microjet compared to the equilibrium flow. The physical justification is found for the diameter of the nozzle, which is the boundary value between the macro- and microjets of vibrationally relaxing gases.

Simulations of astrophysical convection often use different thermal boundary conditions (TBC) at the domain top and bottom, but the consequences of using “mixed” TBC (MTBC) have not been thoroughly examined. Here we show that for MTBC, poorly chosen initial conditions cause a long thermal relaxation which is physically uninteresting and computationally expensive. A couple of mechanisms for avoiding this long relaxation are demonstrated. We also study asymmetries caused by MTBC and find that despite strong asymmetries in the temperature field near the boundaries, MTBC do not meaningfully break the symmetry in the bulk of the flow.

Experiments are performed exploring the dynamics and stability of quasi-one-dimensional lattices of self-propelled millimetric droplets bouncing on a vertically vibrating liquid bath and confined to an annular ring. Depending on the droplet spacing, instability of the lattice may take the form of either out-of-phase oscillations or a striking solitarylike wave. Our results offer a new macroscopic platform to study collective phenomena and emergent structures in the burgeoning field of active and driven matter.

The dynamics of a water droplet impacting on a hydrophobic plate with a cylindrical pore is captured by a nonorthogonal multiple-relaxation-time lattice Boltzmann method. Three distinctive regimes of droplet penetration behaviors: hanging, total crossing, and continuous dripping are found. A phase diagram is constructed for the different regimes, with the phase boundaries determined by the balance among the droplet dynamic pressure, capillary pressure, and viscous drag.

Evaporating a sessile colloidal droplet containing small amounts of salt results in the formation of a ring-shaped stain, resembling the classical coffee-stain effect. However, this ring shape is the only similarity: The structure is actually formed by a single monolayer of particles at the liquid-air interface of the droplet and driven by a solutal Marangoni flow. An experimental study, particle-per-particle, of the formation of this monolayer is presented and numerical simulations are performed to understand the particle aggregation mechanism.

A stationary nonwetting droplet in a constricted tube can attain resonance in response to external seismic stimulation. Key parameters controlling droplet resonance are analyzed through a proposed theoretical model. The nonlinear effect of a fluid system on the resonance is also addressed.

Droplets pushed through small channels can undergo breakup, which limits the throughput in microfluidic applications. We combine theory, numerical simulation, and experiments to explain and quantify the underlying physical breakup process for interacting droplet pairs in a Y-junction. Breakup is driven by spatially varying Young-Laplace pressures, which drive the formation of a neck and a subsequent Rayleigh-Plateau-like pinch-off in the leading droplet. Quantitative analysis suggests that fast interface deformations can dynamically generate Marangoni stresses that significantly impact the dynamics.

In a spinning drop tensiometer, the interfacial tension between two immiscible liquids can be inferred from the equilibrium shape of a drop suspended in a rotating medium. Similarly, an analytical solution is derived for the deformation dynamics of an initially spherical elastic particle suspended in a denser viscous rotating liquid. This gives a proof of concept for a rotating tensiometer for the measurement of the mechanical properties of soft solid beads. Direct numerical simulations are used to validate the theory and identify its limits of applicability.

The transient shape variation and the inner flow field of a rotating liquid droplet are simulated numerically. It is found that two vortices appear in the two-lobed droplet, and a countercurrent flow is formed in between two lobes. The relationship between the maximum deformation and the rotation rate is affected by the surface tension and the density, and the effect of density becomes small as the deformation increases.

The confinement effects in brine, water, and oil are explored in hydrophilic silica nanopores. As the nanopores approach a 1.0-nm radius, the confinement effects are very pronounced, changing the interfacial tension behavior, as well as viscous flow properties. Also, a water-brine layer always stays adsorbed in the silica surface, mediating its interaction with the oil, even hindering the oil infiltration in a case.

Spanwise-harmonic control in a separated flow can modify the mean length of the recirculation region. A second-order sensitivity analysis is used to compute small-amplitude spanwise-harmonic wall controls (blowing/suction or deformation) that most efficiently increase or decrease the mean recirculation length in the flow past a backward-facing step.

Thermoacoustic oscillations in combustion chambers can usually be stabilized with small design changes. Here, an adjoint method is used to identify the required changes and a gradient-based optimization algorithm is used to implement them. The adjoint method reveals the physical mechanisms that are being exploited. These are to dephase the heat release rate from the pressure and to increase the mechanical work done at the acoustic boundaries.

Laminar-flow wings are often seen as an important component of reducing the emissions of future aircraft. However, at high speeds, compressible flow phenomena, such as transonic buffet, limit the flight envelope. The application of scale-resolving simulations is studied to help improve the understanding of the physical mechanisms of this limiting instability.

Dewetting instability of multilayer planar sheets are studied by linear analysis. Several unstable modes are identified, while the maximum growth rate depends on fluid properties. These results provide theoretical guidance to enhance or suppress the dewetting instability via material selection and structure design, enabling fabrication of sophisticated nanostructures for functional fibers and wearable textiles.

The subcritical route to turbulence in shear flows is characterized by metastable localized turbulent-laminar patterns. In plane channel flow, these take the form of intermittent oblique turbulent bands, which either proliferate or decay on timescales that depend on the Reynolds number. A statistical study via direct numerical simulations in a narrow tilted domain leads to the determination of a crossing Reynolds number of around 965, above which the probability for a band to split outpaces its probability to disappear.

The perturbation dynamics over an airfoil laminar separation bubble (LSB) is examined via resolvent analysis. The computational burden of singular value decomposition (SVD) is relieved through the randomized method, and the local energy amplification mechanisms of the LSB are extracted from the global resolvent operator with the discounting approach. With the applications of input and output windows, it is shown that the Kelvin-Helmholtz instability dominates the energy amplification over the LSB, and the optimal momentum-based forcing aligns with the tangential direction of the surface.

Gravity is crucial to the frozen wave instability because it both sets the threshold of applied forcing and limits the growth of unstable perturbations. Through numerical simulations of immiscible liquids in finite containers we examine the transition from a supercritical pitchfork bifurcation in ordinary gravity, which is accompanied by stable finite-amplitude waves, to a degenerate bifurcation in weightlessness, which is followed by large columnar patterns. The vibroequilibria effect plays a symmetry-breaking role in the bifurcation, selecting frozen waves with the heavier fluid displaced upward along the lateral walls.

Lévy walks are abnormal random walks, characterized by Lévy-stable distributed step lengths. Lévy walks occur in a wide range of phenomena such as foraging, socio-physics. etc. An octanoic acid drop performing a Lévy walk on aqueous unsaturated octanoic acid solutions is presented as an inanimate, controlled, and inexpensive model system for studying Lévy walks in a laboratory.

All existing theories of contact lines in fluids assume isothermality. However, an analysis shows that this assumption does not hold for a number of common liquids including water. The variations of temperature are caused by production/consumption of heat on a microscopic scale near the liquid-gas interface due to the effects of compressibility and viscosity.

The propulsion characteristics of Marangoni surfers under confinement is examined. Through experimental measurements and numerical simulations, it is demonstrated that, contrary to what might be the usual expectation, the surfers may propel themselves in the direction of lower surface tension. It is discovered that negative pressure is the primary contributor to the fluid force experienced by the surfer and that this suction force is mainly responsible for the reverse Marangoni propulsion.

Advantages of thermocapillary techniques over other approaches to polymer patterning have been recently recognized and will promote them to further expansion and development. In this perspective, a study of Marangoni instability in a layer of a viscoelastic liquid with a deformable interface is of great importance. Its linear stability analysis shows the presence of (and competition among) the long-wave stationary, short-wave stationary, and oscillatory modes, which leads to the emergence of codimension-two and codimension-three points in the parameter space.

Early time spreading of a water drop on a hydrophilic surface is characterized by the wetted radius which grows linearly in time. We report the unexpected result that the initial spread of surfactant-laden drops is impeded by Marangoni stresses, leading to a large increase in total spreading time. The nonuniform distribution of surfactants at the interface generates Marangoni stresses before the drop-solid contact suppresses film drainage and droplet expansion. Our experiments show that, remarkably, surfactants delay the initial fast motion of the three-phase contact lines.

Three-dimensional direct numerical simulations of the ligaments retraction process over a range of system parameters that account for surfactant solubility, sorption kinetics, and Marangoni stresses are presented. The presence of surfactant inhibits the “end-pinching” mechanism and promotes neck reopening through Marangoni flow induced by the formation of surfactant concentration gradients that drive flow reversal toward the neck.

Free-surface flows of viscous liquid down an inclined plane and past cylinders of various cross-sections are studied theoretically and experimentally. For relatively wide cylinders, a pond of nearly stationary fluid forms upstream of the cylinder, and a dry region without fluid occurs downstream of it. The flow structure in the pond region depends on the cylinder cross-section and curvature at the upstream stagnation point. The theory is used to deduce simplified asymptotic expressions of the force on the cylinders. The work is relevant to volcanic lava flow deflection by barriers.

The lymphatic system relies on a complex arrangement of soft valves to pump lymph. While it is widely assumed that the pumping relies on a fluid-structure interaction mechanism, the role of the valve in this mechanism remains poorly understood. A combination of precision desktop-scale experiments, numerical simulations, and theoretical modeling are used to reveal the relationship between valve geometry and flow properties, which are harnessed to induce flow rectification and pumping. The findings enhance the understanding of directed flow in living systems and open avenues for bio-inspired microfluidic and soft robotic devices

The combination of porosity and permeability of a fluid-loaded structure strongly influences its behavior, but separating the two effects has traditionally proven challenging. Here, results from an experimental and numerical investigation are presented in which the porosity is fixed while systematically varying the permeability of cantilevered strips, which are towed through a viscous fluid. Their steady-state deformations and the associated drags at low and moderate Reynolds ($Re$) numbers are studied. At moderate $Re$, the permeability plays an important role, whereas its effect is small for low $Re$.

The first normal pressure (or stress) difference is directly computed from the local values of the pressure tensor components in molecular dynamics simulations of planar Poiseuille flow for a low molecular weight polymeric fluid. The resulting zero shear rate normal pressure difference agrees very well with the value computed using homogeneous shear simulations and the SLLOD algorithm, and less well with the result of the Coleman-Markowitz equation evaluated at equilibrium. This resolves doubts about the effects of homogeneous thermostats in homogeneous nonequilibrium molecular dynamics simulations.

We conduct resolved simulations of spherical particles suspended in a light fluid in shear flow with a ratio of particle to fluid density in the range of 2.5 to 10 and a focus on inertial rather than gravitational effects. Particle volume fractions range from 5% to 33% and particle Reynolds numbers ($R{e}_p$) of 20 and 40 are examined. Significant effects are found due to the memory of the particles earlier motion related to inertia including on collision rates, momentum transfers, and stress differences. We find that increasing the particle density has similar effects to increasing $R{e}_p$.

The undisturbed flow of a particle controls both the undisturbed flow force and the perturbation (quasi-steady, added-mass and history) forces. With the pairwise interaction extended point particle framework we evaluate the undisturbed flow of each particle in a random array through superposition of the perturbation flow induced by all its neighbors. From this the undisturbed flow statistics can be calculated. We obtain a force consistency relation between drag on an individual particle for use in an Euler-Lagrange simulation and the average drag suitable for use in an Euler-Euler simulation.

The complexity of the problem of clustering of inertial particles in turbulence has intrigued researchers for decades. Here it is shown how a simple one-dimensional flow oscillating at a single frequency explains much of that complexity, and a solution for the Lyapunov exponents of inertial particles subjected to oscillator fluid motion is derived.

We numerically investigate Particle Accumulation Structures (PAS) in noncylindrical liquid bridges (LB) for a high Prandtl number liquid. The work examines the morphological evolution of these structures in weightless conditions for various Marangoni and Stokes numbers, tracer densities, aspect ratios and LB volumes. Additionally, a model is used to interpret the increased ability of concave LBs to support these phenomena over a wider range of the particle Stokes number. The model relies mainly on geometrical factors, i.e., the relationship among the interface curvature, fluid streamline topology, and particle mass effects.

Two new algebraic turbulence models for flows dominated by bubble-induced turbulence (BIT) are presented. The first model, referred to as the algebraic Reynolds normal stress model, is derived from a differential Reynolds stress model for bubbly flows. The second model utilizes one two-equation turbulence model to achieve algebraic expressions for $k$ and $ϵ$ in the BIT dominated cases. If both models are combined, it results in a purely algebraic, explicit relation for the Reynolds normal stresses.

This paper develops a physical model able to describe the dam-break flow of particulate suspensions that sediment progressively during propagation at a constant velocity that solely depends on the mixture properties. The model considers the suspension as an equivalent fluid of constant density and negligible viscosity and leads to a good prediction of the flow duration and deposits shape. These findings allow the formulation of consistent shallow-water equations that can be used to compute the dense basal layer of small-volume pyroclastic flows.

Gas migration through a soft, liquid-saturated granular material involves a strong coupling between the motion of the gas and the deformation of the material. This process is central to many natural and industrial systems, such as the generation and venting of gases from lake beds and waste ponds. Using high-resolution experiments and a simple mechanistic model, grain-scale fluid and solid mechanics are linked with macroscopic migration, trapping, and venting. The largest amount of trapping and the largest venting events are found to occur at intermediate confining stress.

Spectra and statistics in chemically reacting compressible homogeneous isotropic turbulence is studied using numerical simulations. Reaction heat release significantly enhances the spectra of velocity and thermodynamic variables over a wide range of length scales. For exothermal reactions, acoustic mode dominates over the dynamics of compressible velocity and pressure from weak to highly compressible turbulence, and the ratio of compressible to solenoidal kinetic energy and the ratio of compressible to solenoidal dissipation appear to be independent of turbulent Mach number.

It is demonstrated for the first time that by combining Eulerian and Lagrangian flow analyses of tracer particle trajectories in a nearly homogeneous and isotropic turbulence (HIT) in superfluid helium-4 (He II), the turbulence statistics at all length scales can be resolved. The spatial velocity structure functions of the HIT in He II exhibit classical Kolmogorov scaling at large length scales together with anomalous deviations at small scales. Our technique paves the way for future quantitative studies of the similarities and differences between quantum and classical turbulence.

Direct numerical simulations of compressible turbulent channel flows are performed for bulk Mach numbers M${}_b$ = 0.8 and 1.5 and bulk Reynolds numbers Re${}_b$ up to 34000. It is shown that the incompressible and compressible flows become similar at high Re${}_b$. Most of the compressibility effects can be considered using semilocal scaling. Although the streamwise Reynolds stress peak continuously increases with M${}_b$, the change with M${}_b$ decreases at higher Re${}_b$.

The regularity of chirality transfer is numerically investigated in compressible helical turbulent flows. It is found that the compressible component of velocity serves as a medium role for chirality transfer, and the expansion of fluid elements can lead to a stronger lack of mirror symmetry. These conclusions not only uncover the complexity of compressible chiral turbulence but also may serve as a possible controlling method for chiral turbulence.

Pressure loading in massively separated flow regions is intimately tied to the entrainment characteristics of the separated shear layer, which also determines the streamwise distance required for shear layer reattachment. Using stereoscopic particle image velocimetry measurements of a Mach 2.49 longitudinal cylinder wake, a clear relationship between the presence of coherent turbulent structures (i.e., hairpin vortices) and the shear layer reattachment length is demonstrated.

High-Schmidt-number passive scalar fields resulting from an isokinetic release in a turbulent boundary layer are characterized via turbulent scalar length scales, fractal geometry, intermittency, and power spectrum in the inertial-convective (I-C) and viscous-convective (V-C) regimes.The analysis indicates that the V-C scaling behavior deviated significantly from Batchelor’s -1 scaling law for all Reynolds numbers and initial nozzle diameters. The scalar fluctuation intermittency, which generates nonGaussian tails in the PDFs of the scalar fluctuations and produces a large dissipation rate, explains the steep spectral slope in the V-C regime.

Spectra in turbulent channel flow are efficiently represented using a decomposition of the resolvent operator into Orr-Sommerfeld (OS) and Squire (SQ) families of modes. The accuracy of the OS-SQ representation is shown to be superior to the standard resolvent decomposition and results from the ability of the OS and SQ vorticity responses to interact. This interaction is then leveraged to derive Reynolds number scalings for the relative magnitudes of the OS and SQ mode weights for several classes of resolvent modes.

In reacting systems in turbulent fluid environments, a competition exists between the chemical reactions that tend to dump reactant concentration fluctuations and enhance their correlation intensity, and the turbulent mixing that, on the contrary, increases fluctuations and removes relative correlations. It is shown that close to chemical equilibrium, a unique control parameter—the Damkhöler number, based on the reactant Taylor microscale—allows quantitative predictions of the reactants’ statistical properties.

Universality concepts have played a pivotal role in the development of complex systems. This has not been the case in compressible turbulence as no unifying set of parameters has been found to yield universal scaling laws. An analysis supported with DNS databases and studies across the literature shows how universality can indeed be achieved in homogeneous compressible turbulence from specific limiting behavior, paving the way toward a more unified fundamental understanding of compressible turbulence and development of robust compressible turbulence models.

The bifurcations of large-scale jets in a turbulent shear flow driven by a Kolmogorov forcing are studied. These bifurcations, seen also in the mean velocity profile of the forced-dissipative system, are then reproduced using a system at equilibrium, namely, the truncated Euler equations. Statistical properties of the large-scale mode exhibit $1/f$ noise and are in qualitative agreement in both systems.

In recent years, machine-learning techniques have been leveraged to improve closures for the Reynolds-average Navier-Stokes equations. The efficacy of a sparse regression-based method that results in compact, algebraic models and ensures key physical properties such as form invariance is demonstrated. Further, success for learning accurate models from data spanning full high-fidelity sets to experimental (sparse and noisy) data is shown.

A study of the evolution of the wake flow of a square-back Ahmed body is presented. Various ground clearance configurations around the critical case associated with the onset of the lateral bistability are investigated. The oscillation modes vary between stable and bistable states, the corresponding Strouhal numbers for the horizontal (respectively, vertical) evolve from 0.16 (respectively, 0.27) to 0.13 (respectively, 0.18).

Time-harmonic and periodically modulated swirling waves can be excited in an upright circular tank undergoing periodic oscillations in a single degree-of-freedom. The swirling direction is determined by the initial condition. While the time-harmonic swirling waves can reach steady state after the initial buildup, the periodically modulated swirling waves switch back and forth between planar waves and swirling waves.