Turbulent Flow Research Articles

Lagrangian particle tracking at large Reynolds numbers.

In the study of fluid turbulence, the Lagrangian frame of reference represents the most appropriate methodology for investigating transport and mixing. This necessitates the tracking of particles advected by the flow over space and time at high resolution. In the past, the purely spatial counterpart, the Eulerian frame of reference, has been the subject of extensive investigation utilizing hot wire anemometry that employs Taylor's frozen flow hypotheses. Measurements were reported for Taylor scale Reynolds number Rλ > 104 in atmospheric flows, which represent the highest strength of turbulence observed on Earth. The inherent difficulties in accurately tracking particles in turbulent flows have thus far constrained Lagrangian measurements to Taylor scale Reynolds numbers up to approximately Rλ = 103. This study presents the Lagrangian particle tracking setup in the MaxPlanck Variable Density Turbulence Tunnel (VDTT), where Taylor scale Reynolds numbers between 100 and 6000 can be reached. It provides a comprehensive account of the imaging setup within the pressurized facility, the laser illumination, the particles used, and the particle seeding mechanism employed, as well as a detailed description of the experimental procedure. The suitability of KOBO Cellulobeads D-10 particles as tracers within the VDTT is illustrated. The results demonstrate that there is no significant charge exhibited by the particles and that the impact of their inertia on the results is negligible across a wide range of experimental conditions. Typical data are presented, and the challenges and constraints of the experimental approach are discussed in detail.

Autoregressive models for quantification of time-averaging uncertainties in turbulent flows

Autoregressive models for quantification of time-averaging uncertainties in turbulent flows

Impacts of building modifications on the turbulent flow and heat transfer in urban surface boundary layers

This study examines turbulent airflow and upward heat transport in real urban environments using a building-resolving large-eddy simulation model to understand the characteristics of turbulent airflow and upward heat transport when geometrical distributions of buildings are modified. The target areas were two real urban districts within Osaka City, Japan, having different morphological features. In the numerical experiments, the initial condition was set to a neutral condition in which temperature is uniformly distributed vertically, and buildings emitted heat at a constant rate. The results in the two districts indicated that the features of turbulence and heat transport distinctly differed with different building arrangement. Specifically, taller buildings significantly decelerated airflows and induced warming behind buildings. More high-rise buildings (which resulted in a larger building variability) in a district with a larger building density caused a large heat flux and warming at higher levels. The sensitivity experiments in which a density and height variability of buildings were modified showed that a building density at higher levels and a building height variability significantly influenced warming at upper levels. An increased building height variability weakened wind speed and disturbed horizontal heat advection, whereas a large building density caused numerous heat sources.

Influences of blockage ratio and Reynolds number on the spatiotemporal dynamics around a rectangular prism

Particle image velocimetry is used to experimentally investigate the influence of blockage ratio (BR) and Reynolds number (Re) on the turbulent flow around a rectangular prism with depth-to thickness ratio of 3. The prism was selected because it falls within the intermediate regime where the turbulent dynamics is sensitive to the incoming boundary condition. The tested blockage ratios were 2.5%, 5%, and 10% at Reynold numbers of 3000 and 7500. The results are analyzed in terms of the mean flow, turbulent kinetic energy (TKE), frequency spectra, reverse flow area, as well as spectral proper orthogonal decomposition (SPOD). The results indicate that as blockage ratio and/or Reynolds number increase, the tendency of reattachment of the separated shear layer onto the surface of the prism increases while the location of maximum TKE over the prism shifts toward the leading edge, indicating earlier transition of the separated shear layer from laminar to turbulence. For the cases without mean reattachment over the side faces of the prisms, the separated bubble over and downstream of the prism exhibits strong tendency of synchronization in terms of the instantaneous areas of the flow reversal, suggesting a global instability mechanism encompassing the entire prism. In cases with mean flow reattachment, conversely, the low-frequency flapping motion manifests over the prism. SPOD analysis further shows that the relevant shedding dynamics are captured in the first mode and the von Kármán shedding structures have the highest energy.

High-speed turbulent flows towards the exascale: STREAmS-2 porting and performance

Exascale High Performance Computing (HPC) represents a tremendous opportunity to push the boundaries of Computational Fluid Dynamics (CFD), but despite the consolidated trend towards the use of Graphics Processing Units (GPUs), programmability is still an issue. STREAmS-2 (Bernardini et al. Comput. Phys. Commun. 285 (2023) 108644) is a compressible solver for canonical wall-bounded turbulent flows capable of harvesting the potential of NVIDIA GPUs. Here we extend the already available CUDA Fortran backend with a novel HIP backend targeting AMD GPU architectures. The main implementation strategies are discussed along with a novel Python tool that can generate the HIP and CPU code versions allowing developers to focus their attention only on the CUDA Fortran backend. Single GPU performance is analysed focusing on NVIDIA A100 and AMD MI250x cards which are currently at the core of several HPC clusters. The gap between peak GPU performance and STREAmS-2 performance is found to be generally smaller for NVIDIA cards. Roofline analysis allows tracing this behavior to unexpectedly different computational intensities of the same kernel using the two cards. Additional single-GPU comparisons are performed to assess the impact of grid size, number of parallelized loops, thread masking and thread divergence. Parallel performance is measured on the two largest EuroHPC pre-exascale systems, LUMI (AMD GPUs) and Leonardo (NVIDIA GPUs). Strong scalability reveals more than 80% efficiency up to 16 nodes for Leonardo and up to 32 for LUMI. Weak scalability shows an impressive efficiency of over 95% up to the maximum number of nodes tested (256 for LUMI and 512 for Leonardo). This analysis shows that STREAmS-2 is the perfect candidate to fully exploit the power of current pre-exascale HPC systems in Europe, allowing users to simulate flows with over a trillion mesh points, thus reducing the gap between the Reynolds numbers achievable in high-fidelity simulations and those of real engineering applications.

Polymer-doped two-dimensional turbulent flow to study the transition from Newtonian turbulence to elastic instability

Detecting the flow regimes of Newtonian turbulence (NT), elasto-inertial filament (EIF), elasto-inertial turbulence (EIT), and maximum drag reduction (MDR) of polymer solution and their transition has been a hot topic in the last decade. We attempted to detect NT, EIF, EIT, and MDR by visualizing vortex shedding downstream of an array of cylinders that was inserted perpendicular to polymer-doped two-dimensional (2D) flow. Since polymers are stretched at the cylinders, the consequent vortex shedding is affected by viscoelasticity. The flow regimes are characterized based on Weissenberg (Wi) and Reynolds numbers (Re) with the relaxation time of the polymeric solution estimated from capillary-thinning experiments. The flow regimes are observed for different molecular weights of polyethylene oxide and polyacrylamide in solution and are categorized as either vortex type 1, type 2, and type 3 on a Re–Wi map based on flow visualization using particle image velocimetry. In addition, turbulent statistics of these flow regimes are calculated to more fully quantify these flow regimes. We found that vortex types from 1 to 3 have a similarity to NT, EIF, EIT, and MDR. In addition, characteristic turbulent energy transfer without an increase in turbulent energy production was found in the flow regimes of vortex types 2 and 3 of each polymer solution. Our results suggest intriguing parallels between pipe, jet, and 2D turbulent flow for drag-reducing polymeric solutions.

Numerical simulation analysis of turbulent pulsation drag reduction at different intermittent times

The shear stress generated by wall turbulence is the main cause of wall friction resistance in turbulent flow through pipes. This paper investigates the impact of inserting rest periods (regions of constant Reynolds number) within the pulsating operating cycle of velocity on the fully turbulent flow at large time-averaged Reynolds numbers, using the large eddy simulation (LES) method. The study aims to explore the effect of increasing rest periods within pulsations on drag resistance. The dimensionless shear stress and drag reduction rates during different time periods of rest were analyzed. Numerical simulation results indicate that the pulsating velocity operation mode does not necessarily lead to drag reduction; it may even result in increased resistance. Inserting rest periods within the pulsation cycle can achieve drag reduction effects, with the maximum drag reduction rate reaching 21.8%. This paper utilizes large eddy simulation (LES) to compare and validate the feasibility of LES for pipe pulsating operation modes against experimental results and direct numerical simulation (DNS) results from the literature, providing a rationale and proof of concept for further in-depth studies on other pipe pulsating flows.

Oily wastewater treatment using a CNT sorbent in a three-phase circulating fluidized bed

Fine emulsified heavy-oil droplets in wastewater can be easily dispersed in aquatic system, making it challenging to remove them using sorbents. Carbon nanotube (CNT) sorbents were prepared using a m-cresol solvent to remove emulsified heavy oil in wastewater. The bead-shaped and density-controlled sorbents (diameter: 5.22 mm; density: 1169 kg/m3) facilitated efficient contact with emulsified oil and were suitable for use in a three-phase circulating fluidized bed (TPCFB). The sorbents feature a sorption capacity of 2.4 g oil/g sorbent which remained consistent for at least 10 cycles. The hydrodynamic characteristics of the TPCFB reactor (120 × 80 mm; height: 800 mm) were determined and it was found that the circulation of both liquid and solid within the TPCFB was driven by the density difference between the riser and downcomer (Δρ). The solid holdups in the downcomer (εs,d) were almost constant with gas velocity and were proportional to the amount of sorbent inventory. The optimal region for emulsified oil removal was determined to be the downcomer, where turbulent flow was generated that facilitated close contact between the particles and the liquid. The solid circulation rate (Gs) increased to a constant value with Δρ, with the maximum Gs being proportional to the sorbent inventory. The optimal conditions for wastewater treatment were determined to be a maximum εs,d of 0.03 with a Gs value above 10.0 kg/m2s and a sorbent inventory of 90 g. The oil removal rate reached a maximum of 94.8 % at Gs = 12.2 kg/m2s. Used CNT sorbent exhibited unique oil sorption characteristics of CNTs, featuring an even distribution of sulfur on the surface and a weight loss arising from loss of oil at 350–500°C.

Corrosion of Copper, Cupronickel, Nickel Aluminium Bronze and Super-Duplex Stainless Steel Rotating Cylinder Electrodes in Seawater under Turbulent Flow Conditions

The corrosion of uncoated metals in a marine environment restricts the choice of suitable engineering metals and alloys. Anodic dissolution of metal and cathodic reduction of oxygen are important processes and formation of a surface oxide film can affect both electrode reactions. Charge transfer and mass transfer data can provide information on corrosion rate and mechanism. Several metals and alloys having a history of use in marine engineering are considered, including copper, two copper alloys (cupronickel and nickel aluminium bronze) plus a more recent addition to strong, corrosion resistant alloys (a super duplex stainless steel). The rotating cylinder electrode offers the benefits of a controlled, turbulent flow of electrolyte, facilitating controlled mass transfer in a compact cell geometry which is well-suited to bench-top operation. These features are illustrated using a variety of electrochemical techniques, including open-circuit potential vs time monitoring, linear sweep voltammetry, rotation speed step- or potential step current transients and linear polarisation resistance measurements. Seawater taken from Langstone Harbour, Hampshire, UK, was filtered, air-saturated, pH 8.0 and maintained at 25 °C. Dimensionless group correlations and graphical plots (using an analogous approach to the Koutecky-Levich equation for an RDE) enabled contributions of charge transfer-, mixed- and mass transfer- controlled data to be appreciated, allowing the Tafel slopes of electrode reactions, the diffusion coefficient of dissolved oxygen, the mass transfer coefficient for oxygen reduction and the mean corrosion rate to be estimated.

A local measure of the helicity in turbulent flows

Helical structures are a hallmark of hydrodynamic turbulence, and they play a key role in determining the transport and diffusion properties of the flow. Helicity, defined by the linking of velocity and vorticity, is, thus, of primary significance, particularly in systems such as propelled jet wakes. In this paper, we present a novel approach for determining helicity in turbulence, based on one-dimensional filtering methods. First, we test our new technique using analytical solutions and numerical simulations and find that the model can identify high-helicity patterns qualitatively well. Then, we employ the technique in experiments with propelled, helical turbulent flows. The model reliance on single-point velocity measurement may be especially advantageous for measurements with limited acquisition dimensionality.

Effects of a device containing heat pipe sticks on thermal performance of a heat exchanging tube: An experimental study

Applying heat pipes (HP) on heat-exchanging tubes is the main idea of this experimental work. One, two, or three mini copper heat pipes are soldered around a cooper ring to form a new device to passively enhance the performance of a heat-exchanging test tube. The test tube, containing a hot turbulent flow of water-EG solution, is inserted into a cold-water tank. The hot-flow regime is turbulent with Reynolds numbers of 5600 to 10,700. The results for fully developed turbulent flow were compared with existing correlations to validate the experimental setup. The validation showed less than a 10 % difference in both Nusselt number and friction factor. The base fluid in the heat pipes are distilled water and dilute Cu-water nanofluids. The nanofluids with different concentrations (50, 100, 150, and 200 ppm) are applied in HP-sticks. The role of the Cu-nanoparticle's concentration on the performance of the presented device is discussed. The results reveal the notable performance of the device in the target goal. It is reported that the presented device containing HP-sticks filled with water enhances convection rate of the test tube up to 35 %. It is also ascertained that applying the Cu-water nanofluid in the HP-sticks lead to about 14 % better thermal performance, compared to water. The outcomes also reveal that dilute Cu-water nanofluids (φ≤100 ppm) notably improve the device's performance. Conversely, more concentrated nanofluids possibly decrease the device's performance.

Growth characteristics of the mean shear layer in pipe bends with and without a guide vane

The growth characteristics of two identical pipe bends with and without a guide vane are investigated by means of large eddy simulation. The two pipe bends, with a radius of curvature slightly above the separation threshold, are subjected to two fully developed upstream flow conditions, with a corresponding Reynolds number of 11 700 and 24 000. A precursor computation method is employed to provide the fully developed turbulence inflow conditions for all cases. The growth of the mean shear layer in this work is characterized by the local momentum thickness, which measures the extent of momentum deficit confined under the mean shear layer. For both pipe bends, the initial growth of momentum thickness is observed in the first quarter of the bend. The onset location is almost independent of the Reynolds numbers. However, a clear Reynolds number dependence is observed in the onset magnitude, which strongly defines the growth rate thereafter. By examining the mean momentum balance in the bend section, the results show that rather than the adverse pressure gradient, the overall growth characteristics of the mean shear layer, which include the onset location and the growth rate, are better described by the balance between the centrifugal force and the radial pressure gradient. This balance manifests itself as a change in the swirling intensity of the secondary flow. The presence of guide vane significantly suppresses the swirling intensity in the bend section, leading to a noticeable reduction in the overall momentum thickness growth and the production of turbulence in the flow downstream of the bend. Further inspection also indicates that the initial mechanism leading to the suppression of separation at the inner bend is linked to the increasing dominance of the small vortices at the near-wall vicinity relative to the local adverse pressure gradient. Certain aspects pertaining to the turbulent statistics are also discussed.

Including detailed chemistry features in the modeling of emerging low-temperature reactive flows: A review on the application to diluted and MILD combustion systems

Developing and optimizing new reactive systems with carbon-neutral fuels like biofuels, e-fuel, hydrogen, or ammonia is crucial for sustainable energy. This requires advanced technologies capable of fuel flexibility, high efficiency, and minimal pollutant emissions. However, these energy carriers still produce pollutants, especially NOx. To address this, engineers aim to lower combustion process temperatures by adopting different strategies such as burned gas recirculation, staging or increasing the air-to-fuel ratio. Yet, lean flames, though effective at emission reduction, are prone to instability and extinction, posing safety and mechanical risks. Emerging technologies like MILD Combustion, based on burned gas recirculation and reactant dilutions offer interesting solutions. The review article begins by synthesizing experimental studies and numerical simulations of MILD turbulent combustion. It then explores fundamental phenomena specific to diluted combustion (where MILD regimes are included as sub-sets), including autoignition and flame propagation. Using high-fidelity simulations and advanced experiments, it examines flow and mixing roles in reactive zones stabilization. Moving forward, the review paper addresses the inclusion of detailed chemical properties in modeling turbulent combustion systems. Scientific challenges revolve around modeling the intricate interactions between combustion chemistry and flow turbulence while maintaining computational efficiency compatible with industrial constraints. To address this, various simplified chemistry methods – such as reduced, tabulated, or optimized chemistry – have been developed. Additionally, turbulence/chemistry coupling modeling remains unresolved in simulations, with three main routes – geometrical, statistical, or reactor-based approaches – available for turbulent combustion modeling. The state-of-the-art in simplified chemistry and turbulent combustion modeling for low-temperature regimes is then focused on capturing MILD regimes, where there is a crucial impact of dilution by burnt gases, heat transfer, and turbulence mixing on the chemical flame structure. Recent advancements enabled by machine learning and deep learning algorithms are also highlighted. Lastly, the article underscores the critical need for data to validate models, emphasizing the importance of scale-bridging experiments.

Superposition and Amplitude Modulation Effects in Separated Flow

Abstract This article adopts direct numerical simulation method to study concave cavity flow with high Reynolds number, and explores the influence of outer region structure on inner region turbulence of concave cavity flow, as well as the mechanism of coherent structure in the near wall region of concave cavity flow. This article finds that the superposition and amplitude modulation effects of separated flow are weaker than those in smooth wall turbulence. At different positions within the concave cavity, the superposition effect in the reattachment zone is the strongest, while the superposition effect in the separation zone is the weakest, indicating that flow separation prevents large-scale structures from outer region penetrating into the near wall region. The amplitude modulation effect of the reattachment point is the strongest, which is related to the strong adverse pressure gradient of the reattachment point.

Validation of a Central-Difference Finite Volume Method for Analysing Wall Turbulent Heat Transfer Using Transport Equations for Turbulent Heat Fluxes

Abstract The aim of this study is to validate the finite volume method approach using the transport equations for turbulent heat fluxes in turbulent heat transfer at the wall surface. The finite volume method technique is a numerical analysis based on conservation laws of physical quantities in the flow field, with particular emphasis on the conservation of velocity fluctuation intensity in the flow field. The behaviour of turbulent heat flows is analysed in detail using this method and its validity is tested. In the study, heat transfer in channel turbulence is analysed under a constant heat flux heating condition. The individual terms of the temperature variance equation are analysed and the results are shown to be in agreement with a previous study. The residual is small, confirming the high accuracy of the results. The present study analyses the transport equation for the dissipation of the temperature variance. The transport equations for turbulent heat fluxes in the flow direction and in the vertical wall direction are analysed and show good agreement with those of the previous study. The results show that in the vertical wall direction, non-zero terms such as the pressure diffusion term appear to contribute and reduce the accuracy of the transport equations.

Drag modulation by inertial particles in a drag-reduced turbulent channel flow with spanwise wall oscillation

Drag modulation by inertial particles in a drag-reduced turbulent channel flow with spanwise wall oscillation

Limit Cycle Oscillation of Elastic Plate in Supersonic Flow

Fluid–structure interaction of an elastic plate with piezoelectric elements, turbulent freestream flow at Mach 2.5, and a pressurized cavity is investigated computationally and correlated with a recent experiment. The pressure field on the plate is measured using pressure-sensitive paint, and the structural response is observed using the measured voltage of a piezoelectric patch. The measurements show a dominant frequency of oscillation which indicates the likely onset of flutter and a postflutter limit cycle oscillation (LCO). A computational investigation is conducted to study the effects of static pressure differential, temperature differential, cavity pressure coupling, and plate boundary conditions on the linear flutter onset condition and the nonlinear postflutter LCO characteristics. Rivets that connect the plate to the supporting structure are modeled as local constraint in the in-plane direction, and their effect is investigated. Computations show that the coupling between the cavity acoustic and plate structural modes is necessary for flutter onset in the wind-tunnel conditions. Correlation between computed and measured aerodynamic pressure shows reasonable agreement in amplitude and frequency. Computational results are obtained using Piston Theory and also potential flow aerodynamics. Computed and measured pressure LCO mode shapes are extracted and correlated.

Towards bespoke gas permeability by functionally graded structures in laser-based powder bed fusion of metals

Permeable, media transporting, components are an integral part in numerous technical applications. In gas turbines combustors, for example, gaseous oxidizer and fuel are transported separately into the burner, where they are injected and mixed, and subsequently combusted. The mixture homogeneity strongly affects the combustion performance and emissions formation and is, amongst other, determined by the spatial distribution of fuel injection ports. In this context, porous media provide the limiting case for a spatial distribution of media-injecting pores, yet is typically associated with a high pressure drop that yields a loss in efficiency. In this study, possibilities of achieving gas permeability in additively manufactured porous structures are investigated. The objective is to selectively functionalize the permeable layers for gaseous media supply with low pressure loss and, when needed, enable a targeted mixing of different gas streams. For this purpose, a laser-based powder bed fusion process (PBF-LB/M) was used in this study. It offers the opportunity to manufacture varying porosities inside complex monolithic metal parts. To produce the porous structures and to achieve gas permeability, the effect of scan rotation angle, hatch distance, build-up direction and length of the porous specimen is investigated. Due to the high temperatures present in combustion systems, the present work utilizes Inconel 718 material. The AM gas permeable specimen are experimentally characterized by means of surface topography, micro X-ray computed tomography (µXRCT) as well as flow and pressure loss test. The results show, that the AM process parameter provide effective control parameters to adjust the permeability. The strongest effect originates from the hatch distance for a given build-up direction. Depending on the scan rotation, the flow transitions from a turbulent pipe flow to a Darcy flow as present in conventional porous media. A structured alignment and connectivity of pores can be realized as evident in the µXRCT results, surface topography and the flow measurements. Residual powder, powder adhering to the pore walls and stochastic closure of pores or channels lead to deviations and need to be considered when designing respective parts. Nonetheless, the results further show that a directional dependence of the permeability and the build-up direction can be realized and controlled. Consequently, when considering the AM build-strategy in the design of components, this directed permeability can be functionalized in the generation of gas transporting and gas mixing layers separately by adjusting the AM processing parameter.

Schlieren Imaging of Air-coupled Ultrasound

Schlieren imaging uses changes in the refractive index of a medium to permit visualisation of pressure changes in the medium. Schlieren images have been used to observe turbulent convection flow off heat sources, shock waves from bullets, jets and rockets. For a while, in nondestrructive testing it was a popular option to see beam shapes associated with immersion ultrasonic probes as well as pulse-interactions in immersion test conditions. From the perspective of sound transmission, air can be considered another form of a fluid. Although schlieren imaging has been used to illustrate shockwaves in air, the challenge of imaging ultrasound sound in air is significant, primarily due to the extremely high attenuation. This article provides a description of the techniques used to image ultrasound in air using several air-coupled ultrasonic transducers.

Void drift in narrow rectangular channels for highly advective bubbly flows: A macroscopic drift-flux model derived from two-fluid local simulations

This article addresses the issue of reduced models to describe turbulent two-phase flows in industrial applications. This work is an application based on Bois (2021) theoretical derivations to connect the spatially-averaged drift-flux model (DFM) to the local Reynolds-averaged two-fluid model (TFM). It presents new opportunities to calibrate closure laws or propose new models based on finer-scale descriptions. Highly convective steady bubbly flows with wall-peaking void fraction profiles are considered. Dispersed bubbles are weakly deformable. A new model dedicated to void fraction dispersion in flat rectangular channels is developed for application into the DFM. The model is calibrated based on reference local Euler-Euler two-fluid simulations of pressurised water and steam mono-dispersed bubbles, in adiabatic conditions. 64 conditions are considered to cover regularly the 3D parameter space in void fraction, Reynolds and Eötvös numbers. The new formulation proposed shows a very good fit with the post-processed predictions of local CFD simulations along with a substantial reduction in grid requirement. The model contains sub-filter correlations between vapour concentration and mixture or relative velocities. This model derived from local simulations is validated a posteriori in an industrial component scale code where the full space-filtered DFM is implemented and resolved. This work proves the benefits and feasibility of one-by-one model development and calibration based on two-fluid simulations.