Our systems are now restored following recent technical disruption, and we’re working hard to catch up on publishing. We apologise for the inconvenience caused. Find out more: https://www.cambridge.org/universitypress/about-us/news-and-blogs/cambridge-university-press-publishing-update-following-technical-disruption
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
To save content items to your account,
please confirm that you agree to abide by our usage policies.
If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account.
Find out more about saving content to .
To save content items to your Kindle, first ensure coreplatform@cambridge.org
is added to your Approved Personal Document E-mail List under your Personal Document Settings
on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part
of your Kindle email address below.
Find out more about saving to your Kindle.
Note you can select to save to either the @free.kindle.com or @kindle.com variations.
‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi.
‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
Surfactants – molecules and particles that preferentially adsorb to fluid interfaces – play a ubiquitous role in the fluids of industry, of nature and of life. Since most surfactants cannot be seen directly, their behaviour must be inferred from their impact on observed flows, like the buoyant rise of a bubble, or the thickness of a coating film. In so doing, however, a difficulty arises: physically distinct surfactant processes can affect measurable flows in qualitatively identical ways, raising the spectre of confusion or even misinterpretation. This Perspective describes, in one coherent piece, both the equilibrium properties and dynamic processes of surfactants, to better enable the fluid mechanics community to understand, interpret and design surfactant/fluid systems. Specifically, we treat the equilibrium thermodynamics of surfactants at interfaces, including surface pressure, isotherms of soluble and insoluble surfactants and surface dilatational moduli (Gibbs and Marangoni). We describe surfactant dynamics in fluid systems, including surfactant transport and interfacial stress boundary conditions, the competition between surface diffusion, advection and adsorption/desorption, Marangoni stresses and flows and surface-excess rheology. We discuss paradigmatic problems from fluid mechanics that are impacted by surfactants, including translating drops and bubbles, surfactant adsorption to clean and oscillating interfaces; capillary wave damping, thin-film dynamics, foam drainage and the dynamics of particles and probes at surfactant-laden interfaces. Finally, we discuss the additional richness and complexity that frequently arise in ‘real’ surfactants, including phase transitions, phase coexistence and polycrystalline phases within surfactant monolayers, and their impact on non-Newtonian surface rheology.
This article describes recent progress on premixed flame dynamics interacting with acoustic waves. Expressions are derived to determine the stability of combustors with respect to thermoacoustic oscillations. The validity of these expressions is general, but they are illustrated in laminar systems. Laminar burners are commonly used to elucidate the response of premixed flames to incoming flow perturbations, highlight the role of acoustic radiation in their stability, identify modes associated with thermoacoustic intrinsic instabilities and decipher the leading mechanisms in annular systems with multiple injectors. Many industrial devices also operate in a laminar premixed mode such as, for example, domestic gas boilers and heaters equipped with matrix burners for material processing in which unconfined flames are stabilized at one extremity of the system. This article proposes a systematic approach to determine the stability of all these systems with respect to thermoacoustic oscillations by highlighting the key role of the burner impedance and the flame transfer function (FTF). This transfer function links in frequency space incoming flow perturbations to heat release rate disturbances. This concept can be used in the turbulent flame case as well. Weakly nonlinear stability analysis can also easily be conducted by replacing the FTF by a flame describing function in the expressions derived in this work. The response of premixed flames to harmonic mixture compositions and flow-rate perturbations is then revisited and the main parameters controlling the FTF are described. A theoretical framework is finally developed to reduce the system thermoacoustic sensitivity by tailoring the FTF.
We propose an improved adjoint-based method for the reconstruction and prediction of the nonlinear wave field from coarse-resolution measurement data. We adopt the data assimilation framework using an adjoint equation to search for the optimal initial wave field to match the wave field simulation result at later times with the given measurement data. Compared with the conventional approach where the optimised initial surface elevation and velocity potential are independent of each other, our method features an additional constraint to dynamically connect these two control variables based on the dispersion relation of waves. The performance of our new method and the conventional method is assessed with the nonlinear wave data generated from phase-resolved nonlinear wave simulations using the high-order spectral method. We consider a variety of wave steepness and noise levels for the nonlinear irregular waves. It is found that the conventional method tends to overestimate the surface elevation in the high-frequency region and underestimate the velocity potential. In comparison, our new method shows significantly improved performance in the reconstruction and prediction of instantaneous surface elevation, surface velocity potential and high-order wave statistics, including the skewness and kurtosis.
We conduct a well-controlled model experiment for a wide variety of canopy flows. Examples of these include engineering flows such as wind flow, dispersion of scalars through and over urban areas, and the convective heat transfer in many heat exchangers, as well as natural canopies such as flows through terrestrial or aquatic vegetation. We aim to shed the light on fundamental flow and transport phenomena common to these applications. Specifically, the characteristics of mean flow and scalar concentration characteristics of a turbulent boundary layer flow impinging on a canopy, which comprises a cluster of tall obstacles (this can also be interpreted as a porous obstruction). The cluster is created with a group of cylinders of diameter $d$ and height $h$ arranged in a circular patch of diameter $D$. The solidity of the patch/obstruction is defined by $\phi$ (the total planar area covered by cylinders), which is systematically varied ($0.098 \leq \phi \leq 1$) by increasing the number of cylinders in a patch ($N_c$). A point source is placed at ground level upstream of the patch and its transport over and around the patch is examined. Time-averaged velocity and scalar fields, obtained from simultaneous planar particle image velocimetry-planar laser-induced fluorescence (PIV-PLIF) measurements, reveal that the characteristics of wake and flow above porous patches are heavily influenced by $\phi$. In particular, we observe that the horizontal and vertical extent of the wake and scalar concentration downstream of the patches decreases and increases with $\phi$, respectively. Here, the recirculation bubble is shifted closer to the trailing edge (TE) of the patches as $\phi$ increases, limiting the flow from convecting downstream, decreasing the scalar concentration and virtually ‘extending’ the patch in the streamwise direction. As the bubble forms in the TE, vertical bleeding increases and hence the concentration increases above the patch where the cylinders appear to ‘extend’ vertically towards the freestream.
We present a new data reconstruction method with supervised machine learning techniques inspired by super resolution and inbetweening to recover high-resolution turbulent flows from grossly coarse flow data in space and time. For the present machine-learning-based data reconstruction, we use the downsampled skip-connection/multiscale model based on a convolutional neural network, incorporating the multiscale nature of fluid flows into its network structure. As an initial example, the model is applied to the two-dimensional cylinder wake at $Re_D = 100$. The reconstructed flow fields by the present method show great agreement with the reference data obtained by direct numerical simulation. Next, we apply the current model to a two-dimensional decaying homogeneous isotropic turbulence. The machine-learned model is able to track the decaying evolution from spatial and temporal coarse input data. The proposed concept is further applied to a complex turbulent channel flow over a three-dimensional domain at $Re_{\tau }=180$. The present model reconstructs high-resolved turbulent flows from very coarse input data in space, and also reproduces the temporal evolution for appropriately chosen time interval. The dependence on the number of training snapshots and duration between the first and last frames based on a temporal two-point correlation coefficient are also assessed to reveal the capability and robustness of spatio-temporal super resolution reconstruction. These results suggest that the present method can perform a range of flow reconstructions in support of computational and experimental efforts.
The placement of a scaled-down Savonius (drag) vertical-axis wind turbine on model buildings is analysed experimentally by the use of turbine performance and flow field measurements in a wind tunnel. The set-up consists of two surface mounted cubes aligned in the flow direction. The turbine is tested at six different streamwise positions – three on each cube. Velocity field measurements are performed with particle image velocimetry along the centreline of the cubes with and without the turbine. The performance at each position is evaluated based on measurements of the produced torque and the rotational speed of the turbine. It is demonstrated that the common practice of estimating wind resources based on the urban flow field without the turbine present is insufficient. The turbine has a substantial influence on the flow field and thus also on the available power. The performance is found to be optimal in the front and centre of the first building with a significant drop-off to the back. This trend is reversed for the downstream building. Holistically, for more generic geometries and varying wind directions, the results suggest the central position on a building is a good compromise.
We present a new nonlinear mode decomposition method to visualize decomposed flow fields, named the mode decomposing convolutional neural network autoencoder (MD-CNN-AE). The proposed method is applied to a flow around a circular cylinder at the Reynolds number $Re_{D}=100$ as a test case. The flow attributes are mapped into two modes in the latent space and then these two modes are visualized in the physical space. Because the MD-CNN-AEs with nonlinear activation functions show lower reconstruction errors than the proper orthogonal decomposition (POD), the nonlinearity contained in the activation function is considered the key to improving the capability of the model. It is found by applying POD to each field decomposed using the MD-CNN-AE with hyperbolic tangent activation such that a single nonlinear MD-CNN-AE mode contains multiple orthogonal bases, in contrast to the linear methods, i.e. POD and MD-CNN-AE with linear activation. We further assess the proposed MD-CNN-AE by applying it to a transient process of a circular cylinder wake in order to examine its capability for flows containing high-order spatial modes. The present results suggest a great potential for the nonlinear MD-CNN-AE to be used for feature extraction of flow fields in lower dimensions than POD, while retaining interpretable relationships with the conventional POD modes.
The SARS-CoV-2 is transmitted not only through coughing, but also through breathing, speaking or singing. We perform direct numerical simulations of the turbulent transport of potentially infectious aerosols in short conversations, involving repetitive phrases separated by quiescent intervals. We estimate that buoyancy effects due to droplet evaporation are small, and neglect them. A two-way conversation is shown to significantly reduce the aerosol exposure compared with a relative monologue by one person and relative silence of the other. This is because of the ‘cancelling’ effect produced by the two interacting speech jets. Unequal conversation is shown to significantly increase the infection risk to the person who talks less. Interestingly, a small height difference is worse for infection spread, due to reduced interference between the speech jets, than two faces at the same level. For small axial separation, speech jets show large oscillations and reach the other person intermittently. We suggest a range of lateral separations between two people to minimize transmission risk. A realistic estimate of the infection probability is provided by including exposure through the eyes and mouth, in addition to the more common method of using inhaled virions alone. We expect that our results will provide useful inputs to epidemiological models and to disease management.
With the recent rapid development of artificial intelligence (AI) and wide applications in many areas, some fundamental questions in turbulence research can be addressed, such as: ‘Can turbulence be learned by AI? If so, how and why?’ In order to provide answers to these questions, we applied deep learning to the prediction of turbulent heat transfer based only on wall information using data obtained from direct numerical simulations (DNS) of turbulent channel flow. Through this attempt, we investigated whether deep learning could help to improve our understanding of the physics of turbulent heat transfer. Under the assumption that the wall-normal local heat flux can be explicitly expressed through a multilayer nonlinear network in terms of the nearby wall-shear stresses and wall pressure fluctuations, we applied convolutional neural networks (CNNs) to predict the local heat flux. After optimizing the network hyperparameters using a grid searching method, we found that the network can predict the heat flux very accurately with a correlation coefficient of 0.980 between the DNS and the prediction by CNN for the trained Reynolds number, $Re_{\unicode[STIX]{x1D70F}}=180$, and shows similar accuracy at a Reynolds number three times higher than the trained number. This result indicates that relationships between the local heat flux and the nearby inputs are quite insensitive to the Reynolds number within the tested range. In addition, observing the gradient maps of the trained network, we identified the part of the input data that is essential for the local heat flux prediction and the spatial relationship between the local heat flux and the nearby input fields. In addition to obtaining an understanding of the underlying physics, we investigated whether our model could be utilized for turbulence modelling.
In offshore offloading operations, two vessels in a side-by-side configuration experience actions of both ambient water waves and liquid sloshing in internal tanks. Under the excitation of water waves, complex multibody motions are induced, resulting in liquid sloshing in tanks, and concurrently liquid sloshing can feedback to affect the vessels’ motions. The interaction between waves and two barges in a side-by-side configuration coupled with liquid sloshing effects is investigated for a fixed–free arrangement. A numerical model is developed based on the boundary element method to deal with complex wave induced multibody motions coupled with liquid sloshing in internal tanks. Due to the presence of a narrow gap between two vessels, gap resonance may occur, and a damping surface is introduced to suppress an unrealistic response near resonance. Concurrently, physical experiments with and without liquid sloshing effects are carried out. In-depth discussions on motion characteristics are given, and Stokes and non-Stokes natural frequencies associated with liquid sloshing are discussed. The significance of the present study is twofold. Firstly, the experimental measurements provide reference results for validations of numerical simulations. Secondly, this work gives an insight into wave induced motions with liquid sloshing effects under different wave headings which affect vessel operational safety.
The evolution of midwater sediment plumes associated with deep-sea mining activities is investigated in the passive-transport phase using a simplified advection–diffusion-settling model. Key metrics that characterize the extent of plumes are defined based on a concentration threshold. Namely, we consider the volume flux of fluid that ever exceeds a concentration threshold, the furthest distance from and maximum depth below the intrusion where the plume exceeds the threshold, and the instantaneous volume of fluid in excess of the threshold. Formulas are derived for the metrics that provide insight into the parameters that most strongly affect the extent of the plume. The model is applied to a reference deep-sea mining scenario around which key parameters are varied. The results provide some sense of scale for deep-sea mining midwater plumes, but more significantly demonstrate the importance of the parameters that influence the evolution of midwater plumes. The model shows that the discharge mass flow rate and the concentration threshold play an equal and opposite role on setting the extent of the plume. Ambient ocean turbulence and the settling velocity distribution of particles play a lesser yet significant role on setting the extent, and can influence different metrics in opposing ways.
Two models based on convolutional neural networks are trained to predict the two-dimensional instantaneous velocity-fluctuation fields at different wall-normal locations in a turbulent open-channel flow, using the wall-shear-stress components and the wall pressure as inputs. The first model is a fully convolutional neural network (FCN) which directly predicts the fluctuations, while the second one reconstructs the flow fields using a linear combination of orthonormal basis functions, obtained through proper orthogonal decomposition (POD), and is hence named FCN-POD. Both models are trained using data from direct numerical simulations at friction Reynolds numbers $Re_{\tau } = 180$ and 550. Being able to predict the nonlinear interactions in the flow, both models show better predictions than the extended proper orthogonal decomposition (EPOD), which establishes a linear relation between the input and output fields. The performance of the models is compared based on predictions of the instantaneous fluctuation fields, turbulence statistics and power-spectral densities. FCN exhibits the best predictions closer to the wall, whereas FCN-POD provides better predictions at larger wall-normal distances. We also assessed the feasibility of transfer learning for the FCN model, using the model parameters learned from the $Re_{\tau }=180$ dataset to initialize those of the model that is trained on the $Re_{\tau }=550$ dataset. After training the initialized model at the new $Re_{\tau }$, our results indicate the possibility of matching the reference-model performance up to $y^{+}=50$, with $50\,\%$ and $25\,\%$ of the original training data. We expect that these non-intrusive sensing models will play an important role in applications related to closed-loop control of wall-bounded turbulence.
Recent attempts to use deep learning for super-resolution reconstruction of turbulent flows have used supervised learning, which requires paired data for training. This limitation hinders more practical applications of super-resolution reconstruction. Therefore, we present an unsupervised learning model that adopts a cycle-consistent generative adversarial network (CycleGAN) that can be trained with unpaired turbulence data for super-resolution reconstruction. Our model is validated using three examples: (i) recovering the original flow field from filtered data using direct numerical simulation (DNS) of homogeneous isotropic turbulence; (ii) reconstructing full-resolution fields using partially measured data from the DNS of turbulent channel flows; and (iii) generating a DNS-resolution flow field from large-eddy simulation (LES) data for turbulent channel flows. In examples (i) and (ii), for which paired data are available for supervised learning, our unsupervised model demonstrates qualitatively and quantitatively similar performance as that of the best supervised learning model. More importantly, in example (iii), where supervised learning is impossible, our model successfully reconstructs the high-resolution flow field of statistical DNS quality from the LES data. Furthermore, we find that the present model has almost universal applicability to all values of Reynolds numbers within the tested range. This demonstrates that unsupervised learning of turbulence data is indeed possible, opening a new door for the wide application of super-resolution reconstruction of turbulent fields.
We consider the mixing dynamics of an air–liquid system driven by the rotation of a pitched blade turbine (PBT) inside an open, cylindrical tank. To examine the flow and interfacial dynamics, we use a highly parallelised implementation of a hybrid front-tracking/level-set method that employs a domain-decomposition parallelisation strategy. Our numerical technique is designed to capture faithfully complex interfacial deformation, and changes of topology, including interface rupture and dispersed phase coalescence. As shown via transient, a three-dimensional (3-D) LES (large eddy simulation) using a Smagorinsky–Lilly turbulence model, the impeller induces the formation of primary vortices that arise in many idealised rotating flows as well as several secondary vortical structures resembling Kelvin–Helmholtz, vortex breakdown, blade tip vortices and end-wall corner vortices. As the rotation rate increases, a transition to ‘aeration’ is observed when the interface reaches the rotating blades leading to the entrainment of air bubbles into the viscous fluid and the creation of a bubbly, rotating, free surface flow. The mechanisms underlying the aeration transition are probed as are the routes leading to it, which are shown to exhibit a strong dependence on flow history.
The effect of sharp forward-facing steps on boundary-layer transition is systematically investigated in this work in combination with the influence of variations in Mach number, Reynolds number and streamwise pressure gradient. Experiments have been conducted in a quasi-two-dimensional flow at Mach numbers up to 0.77 and chord Reynolds numbers up to 13 million in the Cryogenic Ludwieg-Tube Göttingen. The adopted experimental set-up allows an independent variation of the aforementioned parameters and enables a decoupling of their respective effects on the boundary-layer transition, which has been measured accurately and non-intrusively by means of a temperature-sensitive paint. The functional relations determined between a non-dimensional transition parameter and the non-dimensional step parameters allow the step effect on transition to be isolated from the influence of variations in Mach number, Reynolds number and pressure gradient. Criteria for acceptable heights of forward-facing steps on natural laminar flow surfaces for the examined test conditions are derived from the present functional relations. The measured transition locations are also correlated with the results of linear, local stability analysis for the smooth configuration, enabling the estimation of the step-induced increment of the amplification factor ΔN of Tollmien–Schlichting waves, which can be incorporated in the eN transition prediction method.
The mixing of immiscible oil and water by a pitched blade turbine in a cylindrical vessel is studied numerically. Three-dimensional simulations combined with a hybrid front-tracking/level-set method are employed to capture the complex flow and interfacial dynamics. A large eddy simulation approach, with a Lilly–Smagorinsky model, is employed to simulate the turbulent two-phase dynamics at large Reynolds numbers $Re=1802{-}18\ 026$. The numerical predictions are validated against previous experimental work involving single-drop breakup in a stirred vessel. For small $Re$, the interface is deformed but does not reach the impeller hub, assuming instead the shape of a Newton's Bucket. As the rotating speed increases, the deforming interface attaches to the impeller hub which leads to the formation of long ligaments that subsequently break up into small droplets. For the largest $Re$ studied, the system dynamics becomes extremely complex wherein the creation of ligaments, their breakup and the coalescence of drops occur simultaneously. The simulation outcomes are presented in terms of spatio-temporal evolution of the interface shape and vortical structures. The results of a drop size analysis in terms of the evolution of the number of drops, and their size distribution, is also presented as a parametric function of $Re$.
A novel experiment is presented to study the initial disturbances on a free surface due to the constant acceleration of liquid around a submerged obstacle. The surface response to different obstacle sizes, initial surface heights and fluid velocities is measured using high-speed videography. Perturbations observed on the surface are classified into either jetting or gravity waves by measuring the steepness of growing liquid columns. A classification phase map between these two regimes is obtained and compared with analytical results by Martín Pardo and Nedić (2021). The agreement between decision boundaries is good for high Froude numbers (high fluid velocities) but deteriorates at lower velocities, where viscosity and surface tension effects (not considered in the analytical model) have a greater predominance. The surface profile and perturbation amplitude measured in experiments are also compared against this analytical model. In all cases, the model accurately predicts the corresponding experimental results at the beginning of the motion, but the prediction error increases with time. It is also observed that faster moving surfaces that lead to the onset of jetting have greater prediction accuracies and longer validity times of the predictions.
Tomographic background oriented Schlieren (Tomo-BOS) imaging measures density or temperature fields in three dimensions using multiple camera BOS projections, and is particularly useful for instantaneous flow visualizations of complex fluid dynamics problems. We propose a new method based on physics-informed neural networks (PINNs) to infer the full continuous three-dimensional (3-D) velocity and pressure fields from snapshots of 3-D temperature fields obtained by Tomo-BOS imaging. The PINNs seamlessly integrate the underlying physics of the observed fluid flow and the visualization data, hence enabling the inference of latent quantities using limited experimental data. In this hidden fluid mechanics paradigm, we train the neural network by minimizing a loss function composed of a data mismatch term and residual terms associated with the coupled Navier–Stokes and heat transfer equations. We first quantify the accuracy of the proposed method based on a two-dimensional synthetic data set for buoyancy-driven flow, and subsequently apply it to the Tomo-BOS data set, where we are able to infer the instantaneous velocity and pressure fields of the flow over an espresso cup based only on the temperature field provided by the Tomo-BOS imaging. Moreover, we conduct an independent PIV experiment to validate the PINN inference for the unsteady velocity field at a centre plane. To explain the observed flow physics, we also perform systematic PINN simulations at different Reynolds and Richardson numbers and quantify the variations in velocity and pressure fields. The results in this paper indicate that the proposed deep learning technique can become a promising direction in experimental fluid mechanics.
The zeroth law of turbulence states that, for fixed energy input into large-scale motions, the statistical steady state of a turbulent system is independent of microphysical dissipation properties. This behaviour, which is fundamental to nearly all fluid-like systems from industrial processes to galaxies, occurs because nonlinear processes generate smaller and smaller scales in the flow, until the dissipation – no matter how small – can thermalise the energy input. Using direct numerical simulations and theoretical arguments, we show that in strongly magnetised plasma turbulence such as that recently observed by the Parker Solar Probe spacecraft, the zeroth law is routinely violated. Namely, when such turbulence is ‘imbalanced’ – when the large-scale energy input is dominated by Alfvénic perturbations propagating in one direction (the most common situation in space plasmas) – nonlinear conservation laws imply the existence of a ‘barrier’ at scales near the ion gyroradius. This causes energy to build up over time at large scales. The resulting magnetic-energy spectra bear a strong resemblance to those observed in situ, exhibiting a sharp, steep kinetic transition range above and around the ion-Larmor scale, with flattening at yet smaller scales. The effect thus offers a possible solution to the decade-long puzzle of the position and variability of ion-kinetic spectral breaks in plasma turbulence. The existence of the ‘barrier’ also suggests that, how a plasma is forced at large scales (the imbalance) may have a crucial influence on thermodynamic properties such as the ion-to-electron heating ratio.
The SPARC tokamak is a critical next step towards commercial fusion energy. SPARC is designed as a high-field ($B_0 = 12.2$ T), compact ($R_0 = 1.85$ m, $a = 0.57$ m), superconducting, D-T tokamak with the goal of producing fusion gain $Q>2$ from a magnetically confined fusion plasma for the first time. Currently under design, SPARC will continue the high-field path of the Alcator series of tokamaks, utilizing new magnets based on rare earth barium copper oxide high-temperature superconductors to achieve high performance in a compact device. The goal of $Q>2$ is achievable with conservative physics assumptions ($H_{98,y2} = 0.7$) and, with the nominal assumption of $H_{98,y2} = 1$, SPARC is projected to attain $Q \approx 11$ and $P_{\textrm {fusion}} \approx 140$ MW. SPARC will therefore constitute a unique platform for burning plasma physics research with high density ($\langle n_{e} \rangle \approx 3 \times 10^{20}\ \textrm {m}^{-3}$), high temperature ($\langle T_e \rangle \approx 7$ keV) and high power density ($P_{\textrm {fusion}}/V_{\textrm {plasma}} \approx 7\ \textrm {MW}\,\textrm {m}^{-3}$) relevant to fusion power plants. SPARC's place in the path to commercial fusion energy, its parameters and the current status of SPARC design work are presented. This work also describes the basis for global performance projections and summarizes some of the physics analysis that is presented in greater detail in the companion articles of this collection.