Search

Physics-informed neural networks (PINNs) are widely used to solve forward and inverse problems in fluid mechanics. However, the current PINNs framework faces notable challenges when presented with problems that involve large spatiotemporal domains or high Reynolds numbers, leading to hyper-parameter tuning difficulties and excessively long training times. To overcome these issues and enhance PINNs' efficacy in solving inverse problems, this paper proposes a spatiotemporal parallel physics-informed neural networks (STPINNs) framework that can be deployed simultaneously to multi-central processing units. The STPINNs framework is specially designed for the inverse problems of fluid mechanics by utilizing an overlapping domain decomposition strategy and incorporating Reynolds-averaged Navier–Stokes equations, with eddy viscosity in the output layer of neural networks. The performance of the proposed STPINNs is evaluated on three turbulent cases: the wake flow of a two-dimensional cylinder, homogeneous isotropic decaying turbulence, and the average wake flow of a three-dimensional cylinder. All three turbulent flow cases are successfully reconstructed with sparse observations. The quantitative results along with strong and weak scaling analyses demonstrate that STPINNs can accurately and efficiently solve turbulent flows with comparatively high Reynolds numbers. [ABSTRACT FROM AUTHOR]

The instability mechanism behind a geometrically simple cylindrical roughness element continues to be a challenging topic in fluid mechanics. Considerable progress has been made in understanding the phenomena in recent years, but more research is needed to predict the temporal nature and spatial structure of the dominant instability in a given flow configuration. This is of particular interest, as these instabilities dictate the transition to turbulence and thus are significant for large-scale effects such as skin friction drag. A smoke-flow visualization study with a large variation of parameters, featuring a cylindrical roughness element connected to a linear traverse, has been performed. Results show good agreement with previous investigations and provide further insights into the stability properties, revealing several unexpected effects. For a low roughness aspect ratio $\eta$ , no global instability is detected even at the highest roughness Reynolds number $Re_{kk}$ , whereas a high aspect ratio indicates a delay in the onset of instability. From the acquired visualizations, we constructed the, so far, richest instability diagram of the wake behind an isolated roughness element in the $Re_{kk}\unicode{x2013}\eta$ space, sampled in the same measurement campaign. Furthermore, information regarding the dominant frequency in the wake can be extracted from the visualization images. Our results suggest a new scaling of the frequency as the velocity is increased. Finally, it is shown that the dominant frequency in a certain flow regime can be well predicted using a Strouhal number based on the cylinder diameter and the roughness velocity. [ABSTRACT FROM AUTHOR]

Physics-informed machine-learning (PIML) enables the integration of domain knowledge with machine learning (ML) algorithms, which results in higher data efficiency and more stable predictions. This provides opportunities for augmenting—and even replacing—high-fidelity numerical simulations of complex turbulent flows, which are often expensive due to the requirement of high temporal and spatial resolution. In this review, we (i) provide an introduction and historical perspective of ML methods, in particular neural networks (NN), (ii) examine existing PIML applications to fluid mechanics problems, especially in complex high Reynolds number flows, (iii) demonstrate the utility of PIML techniques through a case study, and (iv) discuss the challenges and opportunities of developing PIML for fluid mechanics. [ABSTRACT FROM AUTHOR]

Turbulence in fluids is an ubiquitous phenomenon, characterized by spontaneous transition of a smooth, laminar flow to rapidly changing, chaotic dynamics. In 1883, Reynolds experimentally demonstrated that, in an initially laminar flow of water, turbulent motions emerge without any measurable external disturbance. To this day, turbulence remains a major unresolved phenomenon in fluid mechanics; in particular, there is a lack of a mathematical model where turbulent dynamics emerge naturally from a laminar flow. Recently, we proposed a new theory of turbulence in gases, according to which turbulent motions are created in an inertial gas flow by the mean field effect of the intermolecular potential. In the current work, we investigate the effect of viscosity in our turbulence model by numerically simulating the air flow at normal conditions in a straight pipe for different values of the Reynolds number. We find that the transition between laminar and turbulent flow in our model occurs, without any deliberate perturbations, as the Reynolds number increases from 2000 to 4000. As the simulated flow becomes turbulent, the decay rate of the time averaged Fourier spectrum of the kinetic energy in our model approaches Kolmogorov's inverse five-thirds law. Both results are consistent with experiments and observations. [ABSTRACT FROM AUTHOR]

The backward-facing step flow (BFSF) is a classical problem in fluid mechanics, hydraulic engineering, and environmental hydraulics. The nature of this flow, consisting of separation and reattachment, makes it a problem worthy of investigation. In this study, divided into two parts, the effect of a cylinder placed downstream of the step on the 2D flow structure was investigated. In Part 1, the classical 2D BFSF was validated by using OpenFOAM. The BFSF characteristics (reattachment, recirculation zone, velocity profile, skin friction coefficient, and pressure coefficient) were validated for a step-height Reynolds number in the range from 75 to 9000, covering both laminar and turbulent flow. The numerical results at different Reynolds numbers of laminar flow and four RANS turbulence models (standard k-ε, RNG k-ε, standard k-ω, and SST k-ω) were found to be in good agreement with the literature data. In laminar flow, the average error between the numerical results and experimental data for velocity profiles and reattachment lengths and the skin friction coefficient were lower than 8.1, 18, and 20%, respectively. In turbulent flow, the standard k-ε was the most accurate model in predicting pressure coefficients, skin friction coefficient, and reattachment length with an average error lower than 20.5, 17.5, and 6%, respectively. In Part 2, the effect on the 2D flow structure of a cylinder placed at different horizontal and vertical locations downstream of the step was investigated. [ABSTRACT FROM AUTHOR]

In 1959 Batchelor predicted that the stationary statistics of passive scalars advected in fluids with small diffusivity k should display a k−1 power spectrum along an inertial range contained in the viscous‐convective range of the fluid model. This prediction has been extensively tested, both experimentally and numerically, and is a core prediction of passive scalar turbulence. In this article we provide a rigorous proof of a version of Batchelor's prediction in the κ→0 limit when the scalar is subjected to a spatially smooth, white‐in‐time stochastic source and is advected by the 2D Navier‐Stokes equations or 3D hyperviscous Navier‐Stokes equations in Td forced by sufficiently regular, nondegenerate stochastic forcing. Although our results hold for fluids at arbitrary Reynolds number, this value is fixed throughout. Our results rely on the quantitative understanding of Lagrangian chaos and passive scalar mixing established in our recent works. Additionally, in the κ→0 limit, we obtain statistically stationary, weak solutions in H−ϵ to the stochastically forced advection problem without diffusivity. These solutions are almost‐surely not locally integrable distributions with nonvanishing average anomalous flux and satisfy the Batchelor spectrum at all sufficiently small scales. We also prove an Onsager‐type criticality result that shows that no such dissipative, weak solutions with a little more regularity can exist. © 2021 Wiley Periodicals LLC. [ABSTRACT FROM AUTHOR]

For this reason we used the fact that the DNS data collapse well for Graph $0 \leqslant r/\eta \leqslant 100$ and drew the black line up to Graph $r/\eta = 100$; for convenience, we denote the range Graph $0 \leqslant r/\eta \leqslant 100$ as the universal dissipative range. Note that difficulties associated with digitising the DNS data result in a large scatter in Graph $Z r$ as Graph $r/\eta$ approaches zero; further, the values of Graph $Z r$ are sufficiently small in this region so as not to affect the calculations. The value of Graph $\bar {\epsilon }$ reported by Kaminsky I et al. i (2020) has been adjusted so that Graph $S 2$ (before normalising by Graph $(\bar {\epsilon } r)^{2/3}$) aligns with the other DNS data in the region Graph $10 \leqslant r/\eta \leqslant 100$. We have used their data of Graph $\bar {\epsilon }$ and Graph $L$; for Graph $Re \lambda = 2297$ we used their value of Graph $\nu = 4.1 \times 10^{-6}$; for Graph $Re \lambda = 7.264\times 10^{4}$ we simply reduced the viscosity to Graph $\nu = 4.1 \times 10^{-9}$, while Graph $\bar {\epsilon }$ and Graph $L$ are unchanged. [Extracted from the article]

The values of Graph $S {z0}$, Graph $\varTheta {z0}$ and the constant viscous dissipation Graph $\varepsilon 0$ are chosen so that the initial density ratio Graph $R {\rho 0}={S {z0}}/{\varTheta {z0}}=4$ and the initial buoyancy Reynolds number Graph $Re {b0}=40$. It is clear on the basis of figure 1 that the difference in Graph $Pr$ is associated with significantly different dependences of diapycnal diffusivities as a function of Graph $Re b$ when the buoyancy Reynolds number is moderate (Graph $Re b). In our model, Graph $\varTheta$ and Graph $S$ are defined in density units so that the equation of state can be written as Graph $\rho =\rho 0+S-\varTheta$. Although the turbulence in the Arctic Ocean is an inherently patchy and intermittent process, as shown in the log-normal distribution of Graph $\varepsilon$ in their figure 7(a), the sampled Graph $Re b$ lies almost entirely within the range of our criterion Graph $0.18. [Extracted from the article]

We combine reduced order modeling and system identification to reconstruct the temporal evolution of large‐scale vortical structures behind the blades of a Rushton impeller. We performed direct numerical simulations at Reynolds number 600 and employed proper orthogonal decomposition (POD) to extract the dominant modes and their temporal coefficients. We then applied the identification algorithm, N4SID, to construct an estimator that captures the relation between the velocity signals at sensor points (input) and the POD coefficients (output). We show that the first pair of modes can be very well reconstructed using the velocity time signal from even a single sensor point. A larger number of points improves accuracy and robustness and also leads to better reconstruction for the second pair of POD modes. Application of the estimator derived at Re = 600 to the flows at Re = 500 and 700 shows that it is robust with respect to changes in operating conditions. [ABSTRACT FROM AUTHOR]

In recent works, we proposed a hypothesis that the turbulence in gases could be produced by particles interacting via a potential and examined the proposed mechanics of turbulence formation in a simple model of two particles for a variety of different potentials. In this work, we use the same hypothesis to develop new fluid mechanics equations that model turbulent gas flow on a macroscopic scale. The main difference between our approach and the conventional formalism is that we avoid replacing the potential interaction between particles with the Boltzmann collision integral. Due to this difference, the velocity moment closure, which we implement for the shear stress and heat flux, relies upon the high Reynolds number condition rather than the Newton law of viscosity and the Fourier law of heat conduction. The resulting system of equations of fluid mechanics differs considerably from the standard Euler and Navier–Stokes equations. A numerical simulation of our system shows that a laminar Bernoulli jet of an argon-like hard sphere gas in a straight pipe rapidly becomes a turbulent flow. The time-averaged Fourier spectra of the kinetic energy of this flow exhibit Kolmogorov's negative five-thirds power decay rate. [ABSTRACT FROM AUTHOR]

The primary objective of this research study was to develop a strategy to examine a key topic of fluid mechanics for undergraduate education by using a theoretical-experimental-computational approach. The key topic examined covered topics from Bernoulli's extended equation, specifically, for calculating head losses in pipelines under turbulent flow conditions. The contribution of the computational approach was appropriate due to the observable reduction in both tedious calculations and optimization time for educational laboratory purposes. The most relevant result was the verification, in a university environment, of trends predicted by theory and experimental results. They correlated in a modified Moody diagram: the Reynolds number, relative roughness, and the Darcy friction factor. In conclusion, the test series showed the advantages and disadvantages of conducting real observations and controlled measurements while performing theoretical analyses. [ABSTRACT FROM AUTHOR]

• • The concept of hydraulic radius can be reinterpreted as a measure of large-scale turbulent flow structure. • • Hydraulic radius as proxy for large-scale eddies enables new framework for bed shear stress model and unified scaling of scour depth. • • Hydraulic radius formulation provides insight into characterizing large-scale flow structure by geometrical boundary conditions. Hydraulic radius has long been customarily used as a characteristic length to assess the cross-sectional geometry and efficiency of channels and conduits. This review paper revisits the concept of the hydraulic radius in light of its extensive applications in recent studies, with the aim of piecing together a state-of-the-art understanding of its physical underpinnings. The original definition of the hydraulic radius is presented as the ratio of cross-sectional flow area to wetted parameter, which inherently integrates the properties of the flow domain and the contacting surface. The use of hydraulic radius as a versatile characteristic length in classical Manning's formula and the Reynolds number underscores its significance in relation to flow resistance. Through a thorough review of its diverse applications, this paper provides a unified understanding of the physical meaning of hydraulic radius, highlighting its interpretation as a measure of large-scale eddies within a given flow domain. This new insight can be justified beyond uniform flows, as evidenced by its use in addressing sediment transport issues in complex vegetated flows, based on turbulence phenomenological theory. Furthermore, the original formulation of the hydraulic radius provides insights into how large-scale flow structure can be quantified with geometrical boundary characteristics, leading to a promising framework for scour prediction based on large-scale vortex size. This paper highlights the versatility and utility of hydraulic radius across various fields of fluid mechanics and sediment transport research, and proposes its potential for future studies as a proxy for large-scale flow structures. [ABSTRACT FROM AUTHOR]

A subject of an experimental fluid mechanics task was measurement of flow around hyperboloid-shaped cooling tower. Experimental part was provided by means of 2D Particle Image Velocimetry (PIV). The contribution contains results from PIV measurements in a range of Reynolds numbers varying from 43000 up to 53000. The tunnels used for the measurement were two different types: Low Speed Wind Tunnel of Eiffel type (LSWT) with low rate of turbulence intensity and Boundary Layer Wind Tunnel (BLWT) with high rate of turbulence intensity. [ABSTRACT FROM AUTHOR]

A re-appraisal of the Richardson’s 1926 dataset [Richardson, L. F. Proc. Roy. Soc. Lond. A 100, 709–737, (1926)] displays an unequivocal non-local scaling for the pair diffusion coefficient, , quite different to the previously assumed locality scaling law , where σl is the pair separation. Consequently, the foundations of turbulent pair diffusion theory are re-examined here and it is shown that pair diffusion is governed by both local and non-local diffusional processess inside the inertial subrange. In the context of generalised energy spectra, E(k) ∼ k− p for 1 < p ≤ 3, the new theory predicts two non-Richardson regimes depending on the size of the inertial subrange: (1) in the limit of asymptotically infinite subrange, non-local scaling laws is obtained, , with γ intermediate between the purely local and the purely non-local scalings, i.e. (1 + p)/2 < γ ≤ 2; and (2) for finite (short) inertial subrange, local scaling laws are obtained, . The theory features a novel mathematical approach expressing the pair diffusion coefficient through a Fourier integral decomposition. [ABSTRACT FROM AUTHOR]

Recently, when studying folding of a SH3 domain, we discovered that the flows of transitions between protein states can be surprisingly similar to turbulent fluid flows. This similarity was not restricted by a vortex pattern of the flow fields but extended to a spatial correlation of flow fluctuations, resulting, in particular, in the structure functions such as in the Kolmogorov theory of homogeneous and isotropic turbulence. Here, we undertake a detailed analysis of spatial distribution of folding flows and their similarity to turbulent fluid flows. Using molecular dynamics simulations, we study folding of another benchmark system—Trp-cage miniprotein, which has different content of secondary structure elements and mechanism of folding. Calculating the probability fluxes of transitions in a three-dimensional space of collective variables, we have found that similar to the SH3 domain, the structure functions of the second and third orders correspond to the Kolmogorov functions. The spatial distributions of the probability fluxes are self-similar with a fractal dimension, and the fractal index decreases toward the native state, indicating that the flow becomes more turbulent as the native state is approached. We also show that the process of folding can be viewed as Brownian diffusion in the space of probability fluxes. The diffusion coefficient plays a role of the key parameter that defines the structures functions, similar to the rate of dissipation of kinetic energy in hydrodynamic turbulence. The obtained results, first, show that the very complex dynamics of protein folding allows a simple characterization in terms of scaling and diffusion of probability fluxes, and, secondly, they suggest that the turbulence phenomena similar to hydrodynamic turbulence are not specific of folding of a particular protein but are common to protein folding. [ABSTRACT FROM AUTHOR]

Intermittencies are commonly observed in fluid mechanics, and particularly, in pipe flows. Initially observed by Reynolds (1883), it took one century for reaching a rather full understanding of this phenomenon whose irregular dynamics (apparently stochastic) puzzled hydrodynamicists for decades. In this brief (non-exhaustive) review, mostly focused on the experimental characterization of this transition between laminar and turbulent regimes, we present some key contributions for evidencing the two concomittant and antagonist processes that are involved in this complex transition and were suggested by Reynolds. It is also shown that a clear explicative model was provided, based on the nonlinear dynamical systems theory, the experimental observations in fluid mechanics only providing an applied example, due to its obvious generic nature. [ABSTRACT FROM AUTHOR]

The Humpback whale tubercles have been studied for more than a decade. Tubercle Leading Edge (TLE) effectively reduces the separation bubble size and helps in delaying stall. They are very effective in case of low Reynolds number flows. The current Computational Fluid Dynamics (CFD) study is on NACA 4415 airfoil, at a Reynolds number 120,000. Two TLE shapes are tested on NACA 4415 airfoil. The tubercle designs implemented on the airfoil are sinusoidal and spherical. A parametric study is also carried out considering three amplitudes (0.025c, 0.05c and 0.075c), the wavelength (0.25c) is fixed. Structured mesh is utilized to generate grid and Transition SST turbulence model is used to capture the flow physics. Results clearly show spherical tubercles outperform sinusoidal tubercles. Furthermore experimental study considering spherical TLE is carried out at Reynolds number 200,000. The experimental results show that spherical TLE improve performance compared to clean airfoil. [ABSTRACT FROM AUTHOR]

A “credible” computational fluid dynamics (CFD) model has the potential to provide a meaningful evaluation of safety in medical devices. One major challenge in establishing “model credibility” is to determine the required degree of similarity between the model and experimental results for the model to be considered sufficiently validated. This study proposes a “threshold-based” validation approach that provides a well-defined acceptance criteria, which is a function of how close the simulation and experimental results are to the safety threshold, for establishing the model validity. The validation criteria developed following the threshold approach is not only a function of Comparison Error, E (which is the difference between experiments and simulations) but also takes in to account the risk to patient safety because of E. The method is applicable for scenarios in which a safety threshold can be clearly defined (e.g., the viscous shear-stress threshold for hemolysis in blood contacting devices). The applicability of the new validation approach was tested on the FDA nozzle geometry. The context of use (COU) was to evaluate if the instantaneous viscous shear stress in the nozzle geometry at Reynolds numbers (Re) of 3500 and 6500 was below the commonly accepted threshold for hemolysis. The CFD results (“S”) of velocity and viscous shear stress were compared with inter-laboratory experimental measurements (“D”). The uncertainties in the CFD and experimental results due to input parameter uncertainties were quantified following the ASME V&V 20 standard. The CFD models for both Re = 3500 and 6500 could not be sufficiently validated by performing a direct comparison between CFD and experimental results using the Student’s t-test. However, following the threshold-based approach, a Student’s t-test comparing |S-D| and |Threshold-S| showed that relative to the threshold, the CFD and experimental datasets for Re = 3500 were statistically similar and the model could be considered sufficiently validated for the COU. However, for Re = 6500, at certain locations where the shear stress is close the hemolysis threshold, the CFD model could not be considered sufficiently validated for the COU. Our analysis showed that the model could be sufficiently validated either by reducing the uncertainties in experiments, simulations, and the threshold or by increasing the sample size for the experiments and simulations. The threshold approach can be applied to all types of computational models and provides an objective way of determining model credibility and for evaluating medical devices. [ABSTRACT FROM AUTHOR]

The mechanical power, total rotor drag, and near-wake velocity of a 1:6 scale model (1.075 m diameter) of the US Department of Energy’s Reference Model vertical-axis cross-flow turbine were measured experimentally in a towing tank, to provide a comprehensive open dataset for validating numerical models. Performance was measured for a range of tip speed ratios and at multiple Reynolds numbers by varying the rotor’s angular velocity and tow carriage speed, respectively. A peak power coefficient CP = 0.37 and rotor drag coefficient CD = 0.84 were observed at a tip speed ratio λ0 = 3.1. A regime of weak linear Re-dependence of the power coefficient was observed above a turbine diameter Reynolds number ReD ≈ 106. The effects of support strut drag on turbine performance were investigated by covering the rotor’s NACA 0021 struts with cylinders. As expected, this modification drastically reduced the rotor power coefficient. Strut drag losses were also measured for the NACA 0021 and cylindrical configurations with the rotor blades removed. For λ = λ0, wake velocity was measured at 1 m (x/D = 0.93) downstream. Mean velocity, turbulence kinetic energy, and mean kinetic energy transport were compared with results from a high solidity turbine acquired with the same test apparatus. Like the high solidity case, mean vertical advection was calculated to be the largest contributor to near-wake recovery. However, overall, lower levels of streamwise wake recovery were calculated for the RM2 case—a consequence of both the relatively low solidity and tapered blades reducing blade tip vortex shedding—responsible for mean vertical advection—and lower levels of turbulence caused by higher operating tip speed ratio and therefore reduced dynamic stall. Datasets, code for processing and visualization, and a CAD model of the turbine have been made publicly available. [ABSTRACT FROM AUTHOR]

We report on a study, employing direct numerical simulations, of the turbulent/nonturbulent interface of a wake in a stably stratified fluid. It is found that thresholds for both enstrophy and potential enstrophy are needed to identify the interface. Using conditional averaging relative to the location of the interface, various quantities of interest are examined. The thickness of the interface is found to scale with the Kolmogorov scale. From an examination of the Ozmidov and Kolmogorov length scales as well as the buoyancy Reynolds number, it is found that the buoyancy Reynolds number decreases and becomes of order 1 near the interface, indicating the suppression of the turbulence there by the stable stratification. Finally the overall rate of loss of energy due to internal wave radiation is found to be comparable to the overall rate of loss due to turbulent kinetic energy dissipation. [ABSTRACT FROM AUTHOR]

Friction factor is an important element in both flow simulations and river engineering. In hydraulics, studies on the friction factor in turbulent regions have been based on the concept of three flow regimes, namely, the fully smooth regime, the fully rough regime, and the transitional regime, since the establishment of the Nikuradze’s chart. However, this study further demonstrates that combining the friction factor with Reynolds number yields a united formula that can scale the entire turbulent region. This formula is derived by investigating the correlation between friction in turbulent pipe flow and its influencing factors, i.e., Reynolds number and relative roughness. In the present study, the formulae of Blasius and Stricklerare modified to rearrange the implicit model of Tao. In addition, we derive a united explicit formula that can compute the friction factor in the entire turbulent regimes based on the asymptotic behavior of the improved Tao’s model. Compared with the reported formulae of Nikuradze, the present formula exhibits higher computational accuracy for the original pipe experiment data of Nikuradze. [ABSTRACT FROM AUTHOR]

One of the major flow phenomena associated with low Reynolds number flow is the formation of separation bubbles on an airfoil’s surface. NACA4415 airfoil is commonly used in wind turbines and UAV applications. The stall characteristics are gradual compared to thin airfoils. The primary criterion set for this work is the capture of laminar separation bubble. Flow is simulated for a Reynolds number of 120,000. The numerical analysis carried out shows the advantages and disadvantages of a few turbulence models. The turbulence models tested were: one equation Spallart Allmars (S-A), two equation SST K-ω, three equation Intermittency (γ) SST, k-kl-ω and finally, the four equation transition γ-Reθ SST. However, the variation in flow physics differs between these turbulence models. Procedure to establish the accuracy of the simulation, in accord with previous experimental results, has been discussed in detail. [ABSTRACT FROM AUTHOR]

Flow structures and downstream wake behind two identical tandem flat plates and a square cylinder has been studied at high Reynolds number and similarities and dissimilarities in coherent and incoherent structures have been identified. It was found that the square cylinder produces higher coherent turbulence kinetic energy (TKE) with peaks up to seven times higher than the flat plates in tandem. Also, turbulence kinetic energy production contours revealed that while the turbulent peak occurs mainly at the edges of the normal flat plates in tandem, it develops at centreline in the wake of the square cylinder. Interestingly, it was observed that the square cylinder exhibits two alternative shedding frequencies along centreline in the wake region. [ABSTRACT FROM AUTHOR]

We study unconfined homogeneous turbulence with a destabilizing background density gradient in the Boussinesq approximation. Starting from initial isotropic turbulence, the buoyancy force induces a transient phase toward a self-similar regime accompanied by a rapid growth of kinetic energy and Reynolds number, along with the development of anisotropic structures in the flow in the direction of gravity. We model this with a two-point statistical approach using an axisymmetric eddy-damped quasi-normal Markovian (EDQNM) closure that includes buoyancy production. The model is able to match direct numerical simulations (DNS) in a parametric study showing the effect of initial Froude number and mixing intensity on the development of the flow. We further improve the model by including the stratification timescale in the characteristic relaxation time for triple correlations in the closure. It permits the computation of the long-term evolution of unstably stratified turbulence at high Reynolds number. This agrees with recent theoretical predictions concerning the self-similar dynamics and brings new insight into the spectral energy distribution and anisotropy of the flow. [ABSTRACT FROM PUBLISHER]

This paper investigates experimentally the correlation and the effect of the turbulence intensity on the pressure drop and the heat transfer mechanism of a heat exchanger with elliptic tubes in a staggered arrangement. The heat exchanger studied in this work is an air–water cross flow heat exchanger with air being the external fluid and water the internal one. The heat exchanger consists of 144 elliptic tubes placed in a staggered arrangement. The experiments were carried out for two setups. The first setup was referring to isothermal conditions for which only air was used, flowing around the tubes. The second setup was referring to non-isothermal conditions with air as the external fluid and water as the internal working fluid, flowing inside the tubes. The Reynolds number for the external air for the isothermal experiments ranged between 3100 and 5200 based on the maximum velocity between the elliptic tubes. The turbulence intensity values varied between 0.9% and 3%. For the non-isothermal measurements the Reynolds number took values from 3100 to 7700 and the turbulent intensity ranged from 0.9% to 3%. The measurements showed that the increase of turbulent intensity led to a decrease in the total pressure drop of the external flow of the heat exchanger together with an enhancement of the heat transfer mechanism. [ABSTRACT FROM AUTHOR]