Bulletin of the American Physical Society
77th Annual Meeting of the Division of Fluid Dynamics
Sunday–Tuesday, November 24–26, 2024; Salt Lake City, Utah
Session C12: Interact: Turbulence |
Hide Abstracts |
Chair: Charles Meneveau, Johns Hopkins University Room: Ballroom A |
Sunday, November 24, 2024 10:50AM - 11:20AM |
C12.00001: INTERACT FLASH TALKS: Turbulence Each Interact Flash Talk will last around 1 minute, followed by around 30 seconds of transition time. |
|
C12.00002: The wind-shade roughness model for turbulent wall-bounded flows Charles Meneveau, Nicholas Hutchins, Daniel Chung To aid in prediction of turbulent boundary layer flows over rough surfaces, a new model is proposed to estimate hydrodynamic roughness based solely on geometric surface information. The model is based on a fluid-mechanics motivated geometric parameter called the wind-shade factor. Sheltering is included using a rapid algorithm adapted from the landscape shadow literature, while local pressure drag is estimated using a piecewise potential flow approximation. Similarly to evaluating traditional surface parameters such as skewness or average slope magnitude, the wind-shade factor is purely geometric and can be evaluated efficiently from knowing the surface elevation map and the mean flow direction. The wind-shade roughness model is applied to over 100 different surfaces available in a public roughness database and some others, and the predicted sandgrain-roughness heights are compared to measured values. Effects of various model ingredients are analyzed, and transitionally rough surfaces are treated by adding a term representing the viscous stress component. |
|
C12.00003: Odd viscosity suppresses intermittency in direct turbulent cascades Sihan Chen, Xander de Wit, Michel Fruchart, Federico Toschi, Vincenzo Vitelli Intermittency refers to the broken self-similarity of turbulent flows caused by anomalous spatio-temporal fluctuations. In this Letter, we ask how intermittency is affected by a non-dissipative viscosity, known as odd viscosity, which appears in parity-breaking fluids such as magnetized polyatomic gases, electron fluids under magnetic field and spinning colloids or grains. Using a combination of Navier-Stokes simulations and theory, we show that intermittency is suppressed by odd viscosity at small scales. This effect is caused by parity-breaking waves, induced by odd viscosity, that break the multiple scale invariances of the Navier-Stokes equations. Building on this insight, we construct a two-channel helical shell model that reproduces the basic phenomenology of turbulent odd-viscous fluids including the suppression of anomalous scaling. Our findings illustrate how a fully developed direct cascade that is entirely self-similar can emerge below a tunable length scale, paving the way for designing turbulent flows with adjustable levels of intermittency. |
|
C12.00004: Time-delay embedding of Lagrangian turbulence to identify precursors of extreme events Rishita Das, Darshna Songara Lagrangian evolution of turbulence dissipation rate and its extreme occurrences may depend on the temporal history of evolution of dissipation rate itself and other flow quantities. This dependence is not well understood and difficult to isolate. In this work, we investigate the latent dynamics of turbulence intermittency present in the time-delay embedding of Lagrangian evolution of energy dissipation rate obtained from direct numerical simulations (DNS) of forced isotropic turbulence. The three-dimensional attractor constructed from its leading eigen-time-delay coordinates, representing the most dominant self-similar features of energy dissipation rate, shows a unique structure in isotropic turbulent flows. To predict the occurrence of extreme events in dissipation rate, we further investigate the time-delay embedding of other turbulence quantities including pressure which is experimentally easily measurable. Delay-embedded chaotic time series is decomposed into linear dynamics and nonlinear (intermittent) forcing using a Hankel alternative view of Koopman (HAVOK) model. Trained on part of a Lagrangian trajectory, the model closely reconstructs the test portion of the trajectory, especially for less intermittent time series. Finally, we demonstrate the use of this intermittent forcing of eigen-time-delay coordinates of pressure and its Hessian as potential precursors for predicting the occurrence of extreme dissipation rate in a turbulent flow. |
|
C12.00005: Universality of intense velocity gradients in different turbulent flows Dhawal Buaria, Alain Pumir A central tenet of turbulence theory is the purported universality of flow properties at small scales. This has been thoroughly investigated and is now nominally well established for mean field quantities, such as the two-point correlation function (velocity spectrum). However, the small scales of turbulence are highly intermittent, as characterized by extreme fluctuations in velocity increments and gradients. In this study, we explore the universality of these extreme fluctuations. To that end, we analyze the statistical properties of velocity gradients from various flows, namely: homogeneous isotropic turbulence; turbulent channel flow at the center of the channel, both from direct numerical simulations, and von Karman mixing tank from laboratory measurements. Our comparison of various unconditional and conditional statistics across these different flows is consistent with universal properties for both mean and extreme events. Notably, the conditional average of strain on vorticity, which is crucial for understanding intermittency and its dependence on the Reynolds number, also shows consistent behavior. Our findings underscore the necessity of directly modeling velocity gradient dynamics as a critical component of turbulence modeling. |
|
C12.00006: Spectra and pseudospectra of mean-linearized turbulent flow over dynamic surfaces Kilian Lapanderie, Ankit Srivastava, Scott T. M. Dawson Carefully designed non-rigid surfaces have the potential to reduce the intensity and drag forces associated with turbulent flow across them. Here, we analyze the mean-linearized stability and energy amplification properties of such systems. This approach is motivated several prior works over the last decade applying resolvent analysis methods to study turbulent flow over surfaces featuring compliance and dynamic wall transpiration. This talk will first show that enabling surface compliance can introduce unstable eigenmodes to the mean-linearized dynamics, even in cases where energy amplification (i.e. resolvent gain) is seemingly suppressed. Next it will be demonstrated that even in cases where such systems are linearly stable, there can be eigenmodes introduced near the marginal stability threshold. We reveal that such eigenmodes can be associated with a counterintuitive nonmodal phenomena where there is a local maximum in the pseudospetra (corresponding to a minimum in the resolvent gain) in the vicinity of this marginally stable eigenvalue. We relate these findings to classical stability analysis results and to a classic counterexample concerning matrix pseudospectra, and further discuss the possible practical implications of these findings. |
|
C12.00007: Turbulent energy and momentum transport across the boundary layer of fractal flexible surfaces AOJIA JIANG, Kourosh Shoele The interaction between the turbulent boundary layer and rough surfaces has attracted the attention of the research community over the past several decades. One natural prototype of such interaction is tree canopies. Yet, little is known about the turbulent interaction with flexible branched tree canopies. Our research uses Large Eddy Simulation (LES) methods to study the turbulent boundary layer interaction with flexible trees, with dynamics represented through a reduced-order Newton-Euler formulation. We will discuss different forms of tree responses and explore the coherent flow structures inside the turbulent boundary layer above and within the tree canopy for different tree flexibilities and branching ratios in two canonical arrangements: aligned and staggered placements. We will then discuss how energy is distributed over the canopy height and identify the spatial and temporal correlations inside the turbulent boundary layer using spectral analysis. |
|
C12.00008: Large-scale structures in compressible turbulence and what they can tell us about dilatational motions Diego A. Donzis, John Panickacheril John, Gregory P Bewley The solenoidal nature of the velocity field in incompressible turbulence imposes strong kinematic constraints on the structure of statistical objects describing the flow. Some of these constraints are relaxed in compressible turbulence due to the existence of dilatational motions. Thus, classical statistical relations derived from these constraints and widely used in incompressible flows (e.g. ratios between longitudinal and transverse integral scales or moments of velocity gradients), are modified for compressible flows. Because these dilatational motions affect the flow, it is important to understand their effect on statistics through those constraints and to assess whether those new relations can provide information on compressiblity effects. We show that this is indeed the case and present some new relations. Using Helmholtz decomposition we show that constraints on both solenoidal and dilatational components redefine the way common statistics should be interpreted. For example, the ratio of longitudinal to transverse integral length scales contains information about isotropy (as in incompressible flows) but also on the amount of dilatation providing a direct way of quantifying ``compressibility'' from easy-to-measure correlations. Furthermore, component correlation functions appear to exhibit universal behavior with their combined effect providing information also about dilatational effects. We support our results with a large database of well resolved DNS. |
|
C12.00009: LES of passive scalar with phase relaxation time in isotropic turbulence Hiromichi Kobayashi, Toshiyuki Gotoh We conduct LES of passive scalar with the phase relaxation time to obtain wider inertial ranges with two -5/3 power-law spectra at large Reynolds numbers. Supersaturation (SS) fluctuations in cloud turbulence are modeled as a passive scalar advected by an isotropic turbulence and found to have two -5/3 power-law ranges with different amplitudes. The condensation-evaporation rate is expressed as the product of the liquid water content and SS in a continuum approximation, which appears as a linear damping term with the phase relaxation time in the equation of the supersaturation. LES model for the present system is developed, and the spectrum and the probability density functions of SS are examined with the focus on the effects of the phase relaxation time. |
|
C12.00010: Effect of subgrid-scale anisotropy on wall-modeled large-eddy simulation for separated turbulent flow Di Zhou, Jane Bae Recent studies have highlighted the pivotal role of subgrid-scale (SGS) model in wall-modeled large-eddy simulation (WMLES) for predicting flow separation. In our investigations of flow over a Gaussian bump using WMLES, we observe that eddy-viscosity-based SGS models often show a nonmonotonic prediction of separation bubble size with grid refinement. In contrast, SGS models augmented with anisotropic stress terms yield more consistent outcomes across different mesh resolutions. To pinpoint where SGS anisotropy becomes crucial, we introduced a virtual interface that divides the computational domain into upstream and downstream sections, applying different SGS models to each. Our findings indicate that just upstream of the bump peak, where a favorable pressure gradient is pronounced, anisotropic SGS stress significantly impacts downstream flow separation. We further conducted detailed budget analyses of the mean momentum and resolved Reynolds stress transport within this region. These analyses reveal that SGS stress plays a crucial role in shaping the boundary layer, influencing the streamwise pressure gradient, and consequently, the formation of downstream separation bubble. Importantly, anisotropic SGS stress positively contributes to the resolved Reynolds stress budget, thereby enhancing the accuracy of predictions. |
|
C12.00011: Spectral Filtering a Model Reactive Tracer; Implications for Large Eddy Simulations Sierra Legare, Marek Stastna This talk will examine the application of several spectral filters to a reactive, passive tracer field from a direct numerical simulation. The tracer reacts according to an idealized non-linear reaction and is subjected to three-dimensional turbulent mixing induced by a Rayleigh-Taylor instability. The mathematical procedure required to obtain the sub-filter-scale advection and reaction quantities from the advection-diffusion-reaction equation governing the reactive tracer evolution will be briefly presented. Each of the sub-filter-scale quantities can be considered in terms of contributions from resolved-scale, cross-scale, and sub-filter-scale interactions. These contributions are quantified for various spectral filter choices and the implications for modelling reactive tracers will be discussed. |
|
C12.00012: Effects of large density contrasts on scale-by-scale energy transfers in Unstably Stratified Homogeneous Turbulence Luminita Danaila, Antoine Briard, Benoît-Joseph Gréa We investigate numerically the effects of significant density contrasts on the dynamics of buoyancy-driven turbulence commonly encountered in geophysical and astrophysical flows. To this aim, we use the Variable Density equations1,2, extending the Boussinesq approximation to large Atwood number flows in the low Mach number limit. In particular, we consider the framework of Unstably Stratified Homogeneous Turbulence (USHT)3, a paradigm of Rayleigh-Taylor mixing layers with a constant background vertical density gradient. |
|
C12.00013: Extension of resolvent framework to turbulent flows laden with low-inertia particles George Papadakis, Stelios Rigopoulos, Rasmus Korslund Schlander We extend the resolvent framework, originally formulated for single-phase turbulent flows, to two-phase flows with low-inertia particles. The particle velocities are modelled using the equilibrium Eulerian model. We analyse the turbulent flow in a vertical pipe with Reynolds number of $5300$ (based on diameter and bulk velocity), for Stokes numbers St+=0-1 and different Froude numbers, as well as the special case where gravity is omitted (1/Frz=0). The resolvent formulation can predict (even with a crude forcing model) the physical phenomena observed in Lagrangian simulations of particulate flows, such as particle clustering and gravitational effects. As for single-phase flows, the operator is low rank over a significant range of wavenumbers and frequencies around the critical layer. When gravity is present, there are two critical layers, one for the velocity field and one for the concentration field. The model also correctly predicts the interaction of near-wall vortices with particle clusters. Overall, the resolvent operator provides a useful framework to explain and interpret many features observed in Lagrangian simulations. Further work will consider the extraction of the operator directly from the Lagrangian particle positions. Success in this endeavour will make the resolvent framework applicable to two-phase flows with high St+ number particles. |
|
C12.00014: Lagrangian finite time- and conditionally-averaged fluctuation relation in isotropic turbulence Hanxun Yao, Tamer A Zaki, Charles Meneveau The entropy generation rate in isotropic turbulence can be defined using local properties from the Kolmogorov-Hill equation, featuring the energy cascade rate as well as the `temperature of turbulence' at a prescribed inertial-range length-scale (Ref. 1). The fluctuation relation (FR) from non-equilibrium thermodynamics that predicts exponential behavior of positive to negative entropy production rate PDFs has been tested using instantaneous flow fields from isotropic turbulence data at Reynolds numbers $Re_\lambda = 1250$ and $Re_\lambda = 433$. We now test the finite-time averaged FR over intervals extending from one to several eddy turnover times, finding that the FR holds and the exponential trend of probability ratios remains valid. Results suggest a minor redefinition of the `temperature of turbulence' that includes a 1/3 factor corresponding to the kinetic energy for each velocity component separately. A key result is that finite-time averaging must be performed within the Lagrangian framework, i.e. integrating along fluid trajectories using the filtered convective fluid velocity. In contrast, the FR fails when using the Eulerian framework, i.e. time-averaging at fixed positions. Finally, we test the single-time FR adherance to the Kolmogorov refined similarity hypothesis, confirming that FR holds (approximately) even when conditioning on different values of locally averaged molecular dissipation rates (Ref. 2). |
|
C12.00015: Wavelet-based resolvent analysis of non-equilibrium channel flow subjected to a sudden transverse pressure gradient Eric Ballouz, Barbara Lopez-Doriga, Scott T. M. Dawson, Jane Bae A better understanding of the influence of non-equilibrium effects on turbulence is important for the analysis of non-stationary turbulent flows, which are ubiquitous in engineering applications. In turbulent channel flow reacting to a strong spanwise pressure gradient imposed suddenly, the Reynolds shear stresses are counter-intuitively attenuated as the flow rotates. In this work, we apply a wavelet-based resolvent analysis framework to study this effect in more detail. Under the wavelet-based formulation, optimal forcing and response modes that encode both time and frequency information are computed for the Navier-Stokes equations linearized about a time-changing mean, Fourier-transformed in the homogeneous directions and wavelet-transformed in time. Prior to the imposition of the spanwise pressure gradient, we find that the optimal forcing exhibits a strong wall-normal component. The mode thus exploits the lift-up mechanism, which couples the wall-normal velocity component with the streamwise and spanwise components. After the onset of the pressure gradient, the wall-normal contribution of the optimal forcing decreases significantly, indicating that lift-up is no longer an efficient way of amplifying velocity perturbations. Finally, by considering the linearized wall-normal velocity-vorticity equations, we argue that it is the misalignment between the mean shear and the mean velocity profiles during the non-equilibrium portion of the flow which attenuates the lift-up term. |
|
C12.00016: Wall-bounded turbulence with logarithmically scale-local interactions Ahmed Elnahhas, Parviz Moin Turbulence can be envisioned as a collection of interacting coherent structures or eddies, that unlike waves, are simultaneously localized in both physical- and scale-space. In the case of the inertial range of homogeneous isotropic turbulence, the energy cascade mechanism can be thought of as a series of successive logarithmically local-in-scale steps in this simultaneous physical- and scale-space representation. This picture has been modeled using dynamical systems with banded interaction kernels, called shell models (L. Biferale Annu. Rev. Fluid Mech. 2003). Analogously, we represent the wall-parallel dimensions by logarithmically spaced representations with local-in-scale interaction kernels while maintaining full physical space representation in the wall-normal direction. It is shown that if the wall-parallel representation is logarithmically spaced waves, the mean velocity profile has the incorrect Kármán constant. It is only when the relative strengths of the terms in the interaction kernel are weighed to mimic wave packets, à la coherent structures, that this local-in-scale representation achieves the mean velocity profile with the correct Kármán constant. We present the statistical states of this dynamical system at Reynolds numbers inaccessible with modern, traditional DNS, since the number of degrees of freedom grows as Reτ2Reτ3/4 as opposed to our (logλReτ)2Reτ3/4, where λ is our chosen logarithmic spacing. |
|
C12.00017: Kinematics of Large Scale Uniform Momentum Zones Ian Jacobi, Guangyao Cui Internal shear layers divide wall-bounded turbulence into distinct uniform momentum zones (UMZs) that provide important insight into the coherent structure and dynamics of the near-wall flow. But traditional techniques for detecting these shear layers and UMZs rely on fixed measurement windows that are too small to capture large-scale motions (LSMs). In this study, we extended a recently proposed moving window approach to UMZ detection in order to identify large-scale UMZs in moderate Reynolds number, wall-bounded flows. We first developed a robust algorithm for the identification of large-scale UMZ boundaries and applied it to streamwise velocity fields from a DNS channel flow. The resulting UMZ boundaries were then classified into linear boundaries (consistent with individual large-scale hairpin packets) and piecewise-linear boundaries (consistent with the superposition of multiple packets). We extracted the size, location, inclination angle, and convective velocities of these UMZ boundaries, and identified correlations between the geometry and kinematical properties of the UMZs. Consistent with our previous work in the spectral domain, we found that the inclination angle of LSMs with respect to the wall is closely connected to their convection velocity. |
|
C12.00018: Mean large-scale linear response of a turbulent flow: scalar transport Paolo Luchini Previous analyses of the mean large-scale linear response of a simulated turbulent channel flow in the plane defined by its two homogeneous directions (turbulent Hele-Shaw flow) are here extended with the inclusion of scalar transport. The same fundamental questions can be asked of the passive scalar (temperature or contaminant concentration, say) as of the main flow's velocity, namely whether the low-frequency and low-wavenumber response function of each quantity exhibits emerging behaviour that can be described by an equivalent large-scale differential equation. This is basically what every turbulence model aims to do, with the distinction that here the relevant information is extracted from a DNS. The extraction technique, already described, e.g., in Luchini, Quadrio & Zuccher, Phys. Fluids 18, 121702 (2006), consists of forcing the system with an externally generated wideband noise-like signal and taking suitable correlations. |
|
C12.00019: Reynolds-number dependence of streamwise velocity variance in wall-bounded turbulent flows Chenning Tong We propose a model for the streamwise velocity variance in wall-bounded turbulent flows. It hypothesizes that the wall-parallel motions of the attached eddies induce internal turbulent boundary layers. The streamwise momentum balance in both outer and inner layers of the internal boundary layers is analysed to obtain the model equations. A logarithmic variance profile is obtained. The peak value of the variance scaled using the friction velocity has a logarithmic dependence on the ratio of the wall-normal length of the flow to the thickness of the internal boundary layer induced by the largest attached eddies, the latter having a dependence on the friction Reynolds number in the form of a Lambert W function. Both the peak and the length ratio are predicted to be unbounded at asymptotically large Reynolds numbers. The model explains the data from the Princeton Superpipe well. The model also predicts that the streamwise velocity fluctuations induced by the attached eddies near the viscous layer scale with the friction velocity; therefore the contributions from the attached eddies to the scaled velocity variance there remains finite at asymptotically large Reynolds numbers. |
|
C12.00020: Data-driven definition of coherent structures in turbulent channels Ricardo Vinuesa, Andres Cremades, Sergio Hoyas The ability of explainable deep learning to detect patterns in turbulent flows has recently been demonstrated. In a recent work (Cremades et al., Nature Communications 15, 3864, 2024), the impact of the Reynolds-stress structures on the evolution of the flow was presented, and their importance for flow predictions was assessed. The present contribution extends the previous results to obtain the importance of every single grid point of a three-dimensional turbulent channel at a friction Reynolds number of 125. First, the evolution of the flow is modeled through deep learning. In particular, a U-net architecture is employed. This deep neural network is able to exploit the existing flow patterns to produce high-quality predictions of the flow (with less than 1% relative error). Then, the importance of each grid point (SHAP value) is evaluated using the expected-gradients method. Finally, new coherent structures based on the impact of each grid point on the prediction of the flow are defined (QSHAP). The new structures are compared against other coherent structures, such as Reynolds-stress structures, streaks, and vortices. The QSHAP structures are strongly related to the Reynolds stresses and streaks in different wall-normal locations. The present methodology opens new opportunities for the analysis of turbulence, and can be extended to the analysis other relevant quantities such as the wall-shear stress or the turbulent kinetic energy. Furthermore, this work paves the way to novel control strategies targeting the most relevant regions of the flow, identified by the SHAP methodology. |
Sunday, November 24, 2024 11:20AM - 12:50PM |
C12.00021: INTERACT DISCUSSION SESSION WITH POSTERS: Turbulence After each Flash Talk has concluded, the Interact session will be followed by interactive poster or e-poster presentations, with plenty of time for one-on-one and small group discussions. |
|
C12.00022: Abstract Withdrawn |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2025 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700