On their way through the blood stream red blood cells need to squeeze through tiny microcapillaries. Physical interactions between the cell membrane and the surrounding flow create a variety of complex cell shapes which may influence the efficiency of oxygen transport. A fascinating example of these is the so-called slipper shape in which the membrane permanently rotates around the liquid core of the red blood cell. Using computer simulations we explain the stability of this shape by a breaking of the fore-aft symmetry in flow.

Continuum mechanics principles are used to obtain the complete linear constitutive equations of a fluid. We obtain the Cauchy stress tensor and entropy that contain additional terms that are missing in the conventional equations. All new terms are due to the relative time rate of change of local temperature. This leads to a new definition of the bulk viscosity. When the results are applied to an ideal gas, the new terms appear to be due to the local time relaxation of intramolecular vibrational energy and when applied to dense gases and liquids they appear to be related to the time relaxation of intermolecular potential energy. Image source: https://stock.adobe.com/images/abstract-background-with-3d-molecules-in-blue-and-red-floating-against-a-blurred-backdrop/919699165

Many particle-settling models overlook the early-active Basset–Boussinesq history force, which can influence the time taken to reach terminal velocity in systems such as marine snow and volcanic ash sedimentation. Our experiments with Stokesian particles highlight the importance of this neglected term, while also revealing the unexpected emergence of a vortex ring, which we model theoretically. These flow structures are important for understanding particle interactions, with implications for both geophysical processes and industrial applications.

Shapley Additive Explanations (SHAP) values are employed to identify high-importance structures. The influence of various coherent structures is assessed in terms of velocity evolution and friction drag generation. While ejections and low-velocity streaks mainly drive the velocity evolution, sweeps emerge as the dominant contributors to friction drag.

A metal disk placed on the water surface normally sinks; however, it has been observed that the disk can remain afloat when a vertical water jet is directed at it from above. The jet displaces water from the disk’s upper surface and, by a mechanism analogous to a hydraulic jump, enlarges the effective immersed volume. The resulting increase in buoyant force offsets the disk’s weight, enabling flotation. A theoretical model based on scaling laws specifies the conditions under which this occurs. Experiments on both flotation and sinking confirm the model’s predictions. A brief video demonstration is available at https://youtu.be/as0wRQj1Zws.

Stenosis developing near arterial bends due to localized plaque buildup narrows the lumen and disrupts the blood flow. This disturbance alters the wall shear stress (WSS) loading and promotes vascular disease progression. In this study, pulsatile flow in straight and bent stenosed artery models was numerically simulated for Reynolds numbers (Re) from 300–1200 and Womersley numbers (Wo) from 7.62–15.24, respectively. Flow imaging at Re = 1200, Wo = 7.62 was performed for validation. Results show that bends intensify flow disturbances, vortex breakdown, and oscillatory shear index, emphasizing the impact of arterial geometry on disease development.

We produced globally unstable (sinusoidal mode) and stable flames (varicose mode) with equivalence ratios of 0.47 and 0.51, respectively. When these flames were subjected to different forcing frequencies and amplitudes, they exhibited heat release rate (HRR) amplitude suppression and resonance. However, under higher detuning conditions, both flames showed amplitude suppression, with the globally unstable flame exhibiting greater suppression. We also found that the HRR amplitude suppression is not uniform throughout the flame and shows spatial variations.

We investigate the impact of bulk viscosity and relaxation model on the shock transition of diatomic gases using the nonlinear coupled constitutive relations, which is an extended model to Navier-Stokes equations for simulating nonequilibrium compressible flows. The findings highlight the distinct mechanisms through which the bulk viscosity and relaxation model improve the shock profiles, offering deeper insight into the physical mechanism inside the shock of diatomic gases.

Evaporation-driven deposition in particle-laden droplets on soft substrates involves complex interactions not yet fully understood. This study presents the first theoretical model coupling droplet dynamics, substrate deformation, and particle transport on a viscoelastic solid. We reveal novel phenomena such as secondary wetting ridge formation, spontaneous symmetry breaking, and an oscillatory contact line instability unique to particle-laden systems. These findings deepen the understanding of deposition patterns and offer insights into the interplay of elasticity, capillarity, and colloidal effects.

How do droplets stay suspended against gravity in an acoustic field, especially when they are too large for the rigid particle assumption to hold? This study explores droplet dynamics beyond the Rayleigh limit, revealing how the balance between acoustic, interfacial, and gravitational forces governs suspension. By modeling the droplet as a fluid, we uncover size-dependent switching between nodes and antinodes, non-monotonic trends in critical energy for suspension, and the potential for droplet sorting, offering new insights into complex droplet behavior in standing acoustic waves.

We investigate the coalescence dynamics of two identical polymer solution drops on a solid substrate. Our primary focus is the influence of the elastocapillary number (Ec), which is defined as the ratio of the sample timescale (i.e., the polymer relaxation timescale) to the experimental viscous timescale of drop merging. The dynamics of the liquid bridge depend non-monotonically on Ec. A combination of surface tension, polymer stress, and viscosity shapes the bridge profile during the process.

A slowly stretched fluid bridge is destined to break, leaving behind a satellite droplet. Combining experiments and one-dimensional simulations, the size of this droplet is shown to be highly reproducible. It is demonstrated to follow a simple formula depending only on the normalized volume of the bridge and the Weber number, so by stretching the bridge faster or increasing its volume the droplet size can be dramatically increased. The robustness of the satellite size is revealed to be due to an underlying universal dynamical solution.

Recalescence is the initial rapid stage of droplet freezing and plays a critical role in the thermal and mechanical dynamics of supercooled water. This study shows that dissolved salts can significantly reduce the recalescence velocity, with up to a 98.8% decrease observed in magnesium chloride solutions. The retardation is attributed to suppressed water diffusivity and a reduced number of active crystal growth sites. A modified Vogel-Fulcher-Tammann equation is incorporated into classical crystal growth theory, yielding a model that accurately captures the retardation across various salt types and concentrations, offering new insights into phase transitions in saline environments.

An electrolyte-filled nanochannel can promote significant thermo-osmotic flow in the presence of wall slip. This flow augments the thermoelectric response of the channel especially in situations with thin electric double layers (relative to the channel width). For example, for a slip length of 40 nm, the thermoelectric power can increase by more than a factor of 200 due to convective charge transport.

We present numerical computations of the dynamic electrophoretic velocity of a charged spherical colloid in unsteady electric fields. Focusing on moderate surface charges and Debye lengths comparable to particle size, we find that when a field is suddenly applied, the electrophoretic mobility initially rises independently of field strength on the momentum diffusion timescale. Over longer times, ion diffusion reshapes the Debye cloud, as the mobility reaches its field-dependent steady state. Under oscillatory fields, the mobility amplitude and phase lag strongly depend on field frequency, which governs the cloud’s ability to adjust within each cycle

We have developed a novel family of self-similar solutions for stress-constrained boundary layer flows over a flat plate. Our work goes beyond previous studies by including power-law stress distributions, creating a complete framework for analyzing boundary layer flows with both increasing and decreasing tangential wall stress. This provides important insights for technologies like magnetic or electric pumping devices.

An experimental study of a dry granular column collapse from a rough slope to a horizontal plane is conducted to explore the dynamic behavior of granular flow and the effect of slope characteristics on the flow dynamics and the final geometry. A computer-vision based recognition method is utilized to analyze the granular profiles. The observed dynamic flow behaviors and the final deposit properties are found to be closely related to the slope and initial granular column configurations. Modified scaling arguments, which consider the aspect ratio, prove to be applicable in this scenario, underscoring the importance of the initial properties of the granular column as a key parameter.

Solute transport in porous media is traditionally modeled with constant dispersivity, assuming Newtonian behavior and uniform viscosity. Using high-resolution micromodel experiments with non-Newtonian shear-thinning fluid, this study demonstrates that such assumptions break down when the local viscosity varies with the shear rate. We reveal a nonmonotonic relationship between dispersivity and flow rate and propose a theoretical model that incorporates shear-dependent viscosity to estimate dispersivity in porous media.

The unsteady separation motion in three-dimensional swept shock-wave/boundary-layer interactions has been shown to exhibit different characteristics compared to nominally two-dimensional interactions. In this numerical study of a sharp-fin induced swept interaction, we show that the separation unsteadiness was strongly correlated with fluctuations in the incoming boundary-layer flow. Oscillations of the separation shock foot occurred at a lower frequency range compared to that of the separation position. Time-periodic forcing of the incoming boundary layer modulated the unsteady separation motion where the response was significantly influenced by the frequency and spanwise form of the forcing.

What kind of flow is sustained in a horizontally rotating cylinder? When the cylinder is completely filled with liquid, the flow tends to solid-body rotation. However, when the cylinder is partially filled, unexpectedly nontrivial vortical structures can arise. This system is both fundamental to science and practical in application, making it of great interest. In this study, we conduct direct numerical simulations to investigate flow structures that emerge in the presence of axial flow and observe a wide variety of flow patterns induced in the system.

Passive flow control strategies for bluff bodies consisting in rigid and flexible appendages has been previously studied, but segmented structures remain less well understood. This study investigates wake modification behind a canonical D-shaped body using arrays of rigid and flexible filaments. Combining PIV and deformation measurements, it is shown that passive 2-D reconfiguration of flexible filaments dominates the reduction of the recirculation bubble, velocity deficit, and drag. The results suggest that wake shaping with flexible or pre-curved rigid appendages offers a promising passive strategy for flow control around blunt bodies

The breaking wave zone of a transitional spot in a Poiseuille flow induces large-scale-outer-layer motions (LSM) that significantly increase the pseudo-turbulent activity. LSM are associated with an intense outer spectral core and a clear scale-separation. The outer spectral core is frozen across the whole layer and the LSM passively penetrates the inner-layer of the spot.

We conduct a study to decipher how the soluble surfactant affects the different unstable temporal modes accountable for thermocapillary instabilities. The H-mode (surface mode), shear mode, and additional two thermocapillary S- and P-modes are identified. We see that the measure of the surfactant solubility parameter stabilizes all the H-mode, S-mode, P-mode, and shear mode instabilities by diminishing their unstable zones. However, the non-dimensional base surface surfactant concentration parameter exhibits both stabilizing and destabilizing roles in the H-mode, S-mode, P-mode, and shear mode instabilities. Moreover, a modal coalescence happens between the H-mode and the S-mode.

The problem of the science behind the impact of droplets on a liquid pool is quite relevant as it uncovers one of the the most abundant fluid interactions. A wave is generated at the impact and it traverses along the air-pool interface. This study provides spatiotemporal resolved whole-field topography of interface following the impact of an immiscible silicon oil droplet on water pool. We report different modes of interaction of the oil droplet based on pool height, droplet viscosity and Weber number, revealing their effect on the surface wave characteristics after the impact.

Using a custom-built three-dimensional scanner, this study reveals how immiscible fluid interfaces destabilize during invasion through disordered porous media. As the flow rate is changed, the interface transitions from stable sheets to tangled, unstable fingers. A stability criterion incorporating pressure dynamics and relative permeability is introduced. The findings bridge two-dimensional theory with complex three-dimensional reality, advancing our understanding of multiscale interactions between flow, structure, and interface dynamics.

How does a beet slice sitting in a thin layer of its own juice develop a translucent fringe? This curious kitchen pattern originates from a dimple (i.e., an indentation in the liquid surface) caused by wetting-induced suction at the beet’s edge. In this study, we show how surface tension, gravity, and viscosity compete to shape this fringe pattern.

Stirring slow viscous flows is a problem common to several physiological systems. We demonstrate that Stokesian oscillatory flows can be stirred via chaotic advection driven by sub-periodic variability and blinking Moffat eddies. Through experiments and simulations in a bifurcating T-section, we identify transitions between open streamlines (non-stirred) and regimes where the stretch-and-fold mechanism, central to Lagrangian chaos, is activated and stirring accentuated. Irreversibly stirring Stokesian flows has been achieved for the first time with greater efficiency than reported in the literature.

The steady Stokes flow around two unequally sized parallel cylinders is solved for exactly using tools from complex analysis and conformal mapping, generalizing the known solutions for a cylinder moving by a plane wall, and two equal cylinders. The resulting flows are highly constrained by the condition that the system be force and torque free, with it impossible to move one cylinder independently of the other.

Microswimmers must adapt their motion to navigate complex and dynamic environments. This study uses reinforcement learning (RL) to train a three-link swimmer to develop adaptive stroke patterns for target-directed navigation. Two learning strategies are developed, optimizing either swimming speed or energy efficiency. Our results reveal that RL not only recovers patterns similar to classical optimization but also adapts to diverse navigation tasks with remarkable flexibility.

This study explores CO2-oil miscible flow in nanochannels using molecular dynamics simulations, revealing how channel material and geometry impact slip length and diffusion. Graphene exhibits a significantly larger slip length (4.5 nm) than silica or MoS2, enhancing oil displacement efficiency. Higher pressures and channel heights improve miscibility, advancing nanoscale fluid dynamics understanding for applications like enhanced oil recovery.

Many-body hydrodynamic interactions play an important role in the dynamics of fluid suspensions. However, there is a dearth of experimental frameworks with which to quantify them. To address this, we develop an optical tweezer-based technique that accurately measures translation-rotation hydrodynamic coupling between trapped colloids with exquisite precision, obtaining a direct reporter of few- to many-body hydrodynamic interactions experimentally. With our technique, we can precisely quantify distant fluid disturbances that are generated by ∼2 pN of hydrodynamic force at 12 particle radii of separation.

Discontinuous shear thickening (DST), extensively studied in dense suspensions, can also occur in dry granular materials when interparticle friction depends on contact force. Using discrete-element simulations, this study shows that introducing a non-Coulombian friction law, mimicking the effect of coatings or lubricants on grain interactions, leads to DST even in the absence of a suspending fluid. A mean-field model is developed that links the microscopic friction law to the bulk rheology, capturing the transition from continuous to discontinuous thickening and offering insights for tuning flow properties in dry industrial powders.

The threshold Froude number associated with instability onset in grain-water mixture flows has not been well defined. In this study, we conduct a temporal stability analysis, perform periodic box numerical simulations, and construct a traveling-wave solution within a debris flow model that incorporates distinct basal friction laws for grains and water. The coupled fluid–particle dynamics reveals an instability onset and subsequent roll wave coarsening behavior which differs from previous findings. The approach is applied to the debris flow event in the Illgraben torrent, Switzerland. The results lead to insights into roll wave initiation and evolution as observed in the field.

This study reveals the onset of a nearly complete segregation regime in submerged granular mixtures rotated in a drum, distinct from dry systems. By introducing an effective density ratio and tracking emergent vortex structures, both experiments and numerical simulations reveal transitions from homogeneous to asymmetrical multi-vortex states. The findings offer predictive capability and mechanistic insights into density-driven segregation in fluid-grain systems.

Many physical systems display symmetries that play a key role in their dynamics. This work introduces a new approach called “symmetry charting” that can be combined with machine learning methods to enforce and exploit symmetries in tasks such as data-driven prediction of future trajectories. As an application, we consider data-driven reduced-order modeling of chaotic dynamics in two-dimensional Kolmogorov flow, which exhibits symmetries under reflections, rotations, and shifts. Symmetry charting enhances prediction of short-time trajectories and long-time statistics, reduces the amount of data required, and effectively simplifies the problem to its most fundamental form.

Larvae of sessile marine species must settle in environments with flow velocities that are orders of magnitude faster than their swimming speed. Previous work has indicated that the roughness of benthic substrates may help to facilitate this difficult process, however, systematic studies of larva-flow-substrate interactions are lacking. We modeled short-range larval transport ($<$10 cm) over a wide range of substrate topographies in wave-like oscillatory flow and found that the transport of larvae to the substrate was aided by recirculatory flow structures generated by millimeter-scale roughness features. Optimal width-to-height ratios were identified that maximized larval settlement.

Forced thermal convection in porous media significantly enhances heat transfer, which is a critical requirement in industrial engineering. However, modeling the heat-fluid-solid coupling in turbulent forced convection necessitates further development. In this study, we develop coupled macroscopic models based on the local thermal nonequilibrium model and the pore-scale prevalence hypothesis. Our results demonstrate that the proposed coupled macroscopic models achieve a high level of accuracy and successfully predict both statistically stationary temperature distributions and overall temperature evolutions.

We examine solute transport and mixing in heterogeneous poroelastic flows under transient forcing. These flows arise in a variety of contexts, ranging from geophysical to biomedical applications, but are not well understood. We show that transport is governed by Lagrangian Coherent Structures (LCS), distinct advective patterns that create transport behaviors that differ markedly from regular flows. These include solute trapping in non-mixing islands, rapid mixing in chaotic zones, and hindered diffusion across transport barriers. These phenomena cannot be resolved via conventional approaches, highlighting the critical role of LCS in shaping solute mixing in transient poroelastic flows.

This study numerically investigates two sources of uncertainties that may influence measurements of flow past a circular cylinder confined in a duct. These being spanwise asymmetries in upstream profiles and measurement plane location. We find variations in upstream profiles strongly affects wake topology. Whereas, due to end effects, uncertainties in measurement plane location strongly influences measurement of flow statistics. Hence, a combination of both may yield uncertainties in flow measurements. These insights may be used to guide experimental investigation of moderate to highly confined flows.

Motivated by previous work on kinetic energy cascades in the ocean, atmosphere, plasmas, and other fluids, we develop a spatio-temporal spectral transfer diagnostic that can be used to study scales of variability in generalized dynamical systems. This spatiotemporal diagnostic can be applied to simulation output or observational data to study the locality of frequency transfers and to calculate triadic interactions in wavenumber and frequency. We test this diagnostic on a simulation of two-dimensional homogeneous isotropic turbulence. We show that the diagnostic is robust to practical problems such as low sampling rates or nonstationarity in time series.

The effects of finite arithmetic precision on direct numerical simulations (DNS) of three-dimensional box turbulence using a spectral method were studied by comparing single and double precision at Taylor-scale Reynolds numbers R${}_{\lambda }$ = 170 and 268. At R${}_{\lambda }$ = 268, significant differences appeared after about 1.6 eddy turnover times, especially in extreme-event statistics such as the maxima and high-order moments of local energy dissipation and enstrophy. In contrast, differences at R${}_{\lambda }$ = 170 were negligible. These results suggest that in high-Reynolds-number DNS, round-off errors from limited precision could significantly affect extreme-event statistics over long simulation times.

In this study, we utilize the acceleration gradient tensor (AGT) as a tool to probe turbulence. It is demonstrated that the AGT principal invariants include not only the spatial characteristics described by the velocity gradient tensor, but also the evolution features such as the temporal evolution of dilatation, strain rate, and rotation rate. Based on the statistical differences in AGT invariants within the viscous sublayer of compressible channel flow, it is confirmed that alternating positive and negative structures near the wall in high-speed flows with strong compressibility induce intense temporal evolution of dilatation.

Moving our measurement frame with the turbulent-nonturbulent interface extends observation of local scalar and velocity structures and enables quantification of two fields: the finite-time Lyapunov field, which marks advective barriers (engulfment), and the diffusive-barrier field, which marks viscous barriers (nibbling). For integration times below the integral scale, conditional averages show both fields correlate with interfacial layers. Over longer times, conditionally averaged diffusive momentum flux demonstrates turbulent growth into the irrotational domain. Diffusive flux parallel to the interface concentrates in a superlayer comparable to the Taylor microscale.

Predicting drag over rough surfaces remains a major challenge in turbulent flow research due to the complexity and diversity of roughness topographies. We show that a deep convolutional autoencoder can distill O(100) disparate experimental and direct numerical simulation (DNS) roughness maps to just three latent space variables, which a small feedforward neural network converts directly into equivalent sand-grain roughness. These ultra-low-dimensional representations can be utilized to accurately predict sand-grain roughness height for completely new surfaces generated via the decoder network, charting a practical, generative and data-driven path towards generalizable rough-wall modeling.

Despite extensive efforts, the mechanisms of drag reduction via transverse wall forcing are not fully understood; here, we emphasize the importance of the Stokes strain rate (SSR). Imposed through a streamwise-periodic square-wave type forcing, SSR forcing is found to be strong and impulsive during reversal of forcing direction but nearly zero otherwise. As the wavelength extends beyond optimal conditions, the impulsive SSR-topology promotes the asymmetry between a short turbulence attenuation (high SSR), and the extended recovery phase (near-zero SSR), while the skin-friction is marked by an out-of-phase response. These insights may also prove valuable in passive forcing surrogates.

Understanding plastron deformation of hydrophobic surfaces under turbulent flow is crucial for optimizing drag reduction. This study experimentally reveals that longitudinal plastron deformation significantly undermines the drag reduction performance of micro-grooved hydrophobic surfaces. By correlating changes in plastron morphology with skin friction and turbulence characteristics, we demonstrate that reducing groove length effectively suppresses plastron longitudinal deformation and enhances drag reduction. These findings provide valuable insights for the design of hydrophobic surfaces for turbulent drag reduction.

The compression or expansion of turbulent mixing layers often occurs under anisotropic strain rates, causing the mixing layer to compress or expand at different rates in different directions. Anisotropic strain patterns are common in nozzle flows and implosions yet are rarely addressed in turbulence models. The treatment of the turbulent length scale under bulk compression is investigated for a two-equation Reynolds—Averaged Navier—Stokes model. Comparing to strained implicit large eddy simulations, the mixing layer properties are better predicted when accounting for the alignment of the strain rates and the mixing layer growth direction.

Weak Euler solutions have been hypothesized to explain the d’Alembert paradox of non-vanishing drag. A difficulty is the “weak-strong uniqueness” property, which requires that an admissible weak Euler solution must coincide with the smooth Euler solution for the same initial data. Using the Josephson-Anderson relation adapted from superfluids, we show that weak-strong uniqueness for d’Alembert’s solution requires mild conditions. To explain drag we therefore predict that these conditions are violated by violent eruption of very thin boundary layers. We discuss observational signatures and explain how the small length-scales involved could threaten the validity of a hydrodynamic description.

Two distinct hypotheses, related to upwind and turbine scales, for the formation of wake meandering —the large-scale, periodic oscillating motion of the far wake of a wind turbine —are tested via LES by varying the incoming scales. Wake meandering is observed throughout the range of scales but exhibits variation in turbulence and spectral characteristics. Two wake meandering scales are observed: (1) turbine influenced scale with regular Strouhal number, St ≈ 0.3 and corresponding harmonics, with a similar wake center distribution across all upwind conditions, and (2) upwind produced scale if low upwind scales are present, St < 0.1 with irregular wake center distributions.

This study investigates the structure and dynamics of the quiescent momentum core (QMC) and introduces its thermal counterpart—the quiescent thermal core (QTC)—in compressible channel flows using direct numerical simulations at moderate to high Reynolds and Mach numbers. While the QMC retains features observed in incompressible flows, such as low turbulence and sharp interfacial layers, the newly identified QTC displays distinct thermal characteristics, including enhanced thickness and strong temperature gradients. These findings clarify how compressibility modifies core region dynamics and energy transport, extending the concept of core-layer structures to thermal fields in high-speed turbulent flows.

We study the small-scale anisotropy of a passive scalar field in homogeneous isotropic turbulence with a uniform mean scalar gradient. We extend linear response theory of turbulence to predict how the scalar gradient induces anisotropy, focusing on second- and third-order mixed velocity-scalar structure functions. The theoretical predictions generally agree with results by direct numerical simulation.

Bayesian approaches to olfactory search, the problem of tracking a source of passive scalars in a turbulent flow, have been very successful but remain understudied in realistic, correlated flows. By searching in realistic direct numerical simulations data, we study the effects of correlations (i.e., structure in the scalar field) on the performance of Bayesian strategies. We find that short-range correlations impede performance but can be mitigated with prior knowledge and additional optimization; meanwhile, large-scale structure typical of windy flows helps the agent reach the target faster, without need for prior knowledge. We also argue that there are optimal values for the observation rate and threshold for detection.

Below critical temperature, turbulence prevents the spontaneous phase separation of binary mixtures, resulting in a phase arrested state of emulsion. The current study explores if a Kolmogorov-like energy cascade exists in fully developed binary fluid turbulence. Using exact relations and direct numerical simulations with up to 1024${}^3$ grid points, we show that the combined kinetic and interfacial energy exhibits a cascade with a constant transfer rate across the inertial scales. In addition, the cascade rates computed from the three exact laws in divergence, alternative and correlator forms show excellent agreement, thus confirming the equivalence between the three formulations.

Despite decades of research on compressible turbulent channel flows (CTCFs), studies inconsistently employ either spatially uniform or density-weighted body forces. We conduct direct numerical simulations of CTCFs with symmetric (cold) and asymmetric (cold/quasi-adiabatic) thermal walls to systematically assess the impact of these two forcing strategies on turbulence statistics. While differences are minimal in symmetric cases, strong compressibility or large wall temperature differences lead to notable discrepancies, highlighting the importance of force selection in high-Mach or thermally asymmetric flows.

In this study, we perform a direct numerical simulation of a turbulent plane Couette flow over a two-dimensional rod-roughened wall to examine the effect of surface roughness on this type of flow. We find that the surface roughness causes the decreased turbulent activity in the outer layer due to the weakening of the large-scale negative u’-component. An analysis of the large-scale turbulent structures in the outer layer shows that the reduction of the streamwise coherence of the large-scale u’-structure with less energy in the outer layer is due to the suppressed development of a hairpin packet resulting from the weakened roll-cell motions with less influence on the near-wall region.

Successive nucleation of a large number of vortices is observed in the wake of the impenetrable barrier accompanied by the emergence of sound waves. A coarsening stage follows where vortices interact with a portion of them being annihilated eventually yielding turbulent response. Depending on the strength and velocity of the stirring potential, different vortex configurations emerge, ranging from vortex dipoles, to vortex clusters, and randomly distributed vortex-antivortex pairs. The dipole and clustered configurations exhibit Kolmogorov-like scaling in the incompressible kinetic energy spectrum, whereas the random vortex-antivortex configurations are characterized by Vinen-like scaling.

In this study, with the prediction directly from the framework based on the Zakharov equation using kernels in Hamiltonian form, instability regions were obtained in different water depths. Extensive numerical simulations were conducted using a higher-order spectral method to investigate the modulational instability. Numerical simulations confirm that sidebands grow exponentially, suggesting that there is noticeable modulational instability in shallow water at dimensionless water depth 0.8. An evident amplification can also be achieved.

This study studies the generation of extreme waves in a physical wave flume and in a numerical wave tank based on the higher order Peregrine solutions to the Schrodinger equation. Higher harmonics of the wave elevations during modulation and demodulation are extracted and analyzed. Through spectral analysis, the nonlinear energy transfer characteristics of Peregrine solutions are identified.

This paper introduces inverse reinforcement learning to discover objectives in fish schooling. The methodology is not specific to fish schools and applicable across other natural systems. It provides a new path to bioinspired optimization by analyzing data to infer goals rather than a-priori specifying them.

The unsteady load caused by gusts (time-varying atmospheric wind disturbances) is an essential factor affecting flight safety. To mitigate gust loads, various flow control methods have been used. However, these methods are effective only in limited ranges, and a full theory of gust load reduction has been elusive. Here we propose a theoretical method for gust load reduction which we have validated through experiments for a certain range of gust amplitudes and frequencies. Our results show that the method reduces lift fluctuations by up to 90%, and provides insight into the relevant flow mechanism.

Physical systems are often described by partial differential equations. In multidimensional, time-dependent cases, these equations form distinct regions with different dynamics that can be clustered. We present a new hybrid method for flow clustering by describing each sample point using equation-based features. The method works in both Eulerian and Lagrangian frameworks. Our results in Lagrangian framwork show how the clusters shift dynamically over time when applied to transient or turbulent data.