Separated flow structures around a cylindrical obstacle in a narrow channel

A detailed experimental study is performed on the separated flow structures around a low aspect-ratio circular cylinder (pin-fin) in a practical configuration of liquid cooling channel. Distinctive features of the present arrangement are the confinement of the cylinder at both ends, water flow at low Reynolds numbers (Re = 800, 1800, 2800), very high core flow turbulence and undeveloped boundary layers at the position of the obstacle. The horseshoe vortex system at the junctions between the cylinder and the confining walls and the near wake region behind the obstacle are deeply investigated by means of Particle Image Velocimetry (PIV). Upstream of the cylinder, the horseshoe vortex system turns out to be perturbed by vorticity bursts from the incoming boundary layers, leading to aperiodical vortex oscillations at Re = 800 or to break-away and secondary vorticity eruptions at the higher Reynolds numbers. The flow structures in the near wake show a complex three-dimensional behaviour associated with a peculiar mechanism of spanwise mass transport. High levels of free-stream turbulence trigger an early instabilization of the shear layers and strong Bloor–Gerrard vortices are observed even at Re = 800. Coalescence of these vortices and intense spanwise flow inhibit the alternate primary vortex shedding for time periods whose length and frequency increase as the Reynolds number is reduced. The inhibition of alternate vortex shedding for long time periods is finally related to the very large wake characteristic lengths and to the low velocity fluctuations observed especially at the lowest Reynolds number.     

PIV–LES analysis of channel flow rotating about the streamwise axis

The flow field of a channel rotating about the streamwise axis is analyzed experimentally and numerically. The current investigations were carried out at a bulk velocity based Reynolds number of Re_m = 2850 and a friction velocity based Reynolds number of Re_τ = 180, respectively. Particle-image velocimetry (PIV) measurements are compared with large-eddy simulation data to show earlier direct numerical simulation findings to generate too large a reverse flow region in the center region of the spanwise flow. The development of the mean spanwise velocity distribution and the influence of the rotation on the turbulent properties, i.e., the Reynolds stresses and the two-point correlations of the flow, are confirmed in both investigations. The rotation primarily influences those components of the Reynolds shear stresses, which contain the spanwise velocity component. The size of the correlation areas and thus the length scales of the flow generally grow in all three coordinate directions leading to longer structures. Furthermore, experimental results of the same channel flow at a significantly lower bulk Reynolds number of Re_{m, l} = 665, i.e., a laminar flow in a non-rotating channel, are introduced. The experiments show the low Reynolds number flow to become turbulent under rotation and to develop the same characteristics as the high Reynolds number flow.     

3D-PTV measurements in a plane Couette flow

Genuine plane Couette flow is hard to realize experimentally, and no applications of modern spatially resolving measurement techniques have been reported for this flow so far. In order to resolve this shortcoming, we designed and built a new experimental facility and present our first results here. Our setup enables us to access the flow via 3D particle tracking velocimetry and therefore to obtain truly three-dimensional flow fields for the first time experimentally in plane Couette flow. Results are analyzed in terms of basic flow properties, and a clear distinction of flow regimes (laminar for Re < 320, transitional for 320 < Re < 400, and turbulent when Re > 400) could be made. Comparison with DNS data shows good agreement in the turbulent regime and builds trust in our data. Furthermore, vortical coherent structures are studied in detail with the additional help of kalliroscope imaging, and the typical vortex spacing is determined to be roughly one gap width. As a noteworthy result, we find that the onset of the turbulent regime coincides with the range of Reynolds numbers at which a distance of 100 wall units is comparable to the gap width.     

Finite element solution of the streamfunction-vorticity equations for incompressible two-dimensional flows

The streamfunction-vorticity equations for incompressible two-dimensional flows are uncoupled and solved in sequence by the finite element method. The vorticity at no-slip boundaries is evaluated in the framework of the streamfunction equation. The resulting scheme achieves convergence, even for very high values of the Reynolds number, without the traditional need for upwinding. The stability and accuracy of the approach are demonstrated by the solution of two well-known benchmark problems: flow in a lid-driven cavity at Re ? 10,000 and flow over a backward-facing step at Re = 800.     

ADAPTIVE GRID REFINEMENT USING CELL-LEVEL AND GLOBAL IMBALANCES

A methodology for local solution-adaptive mesh refinement in computational fluid dynamics (CFD) using cell-level and global kinetic energy balances is formulated and tested. Results are presented for two two-dimensional steady incompressible laminar benchmark problems: a lid-driven cavity (Reynolds number Re=1000) and a backward-facing step (Re=400). It is demonstrated that local kinetic energy imbalance correlates with local solution accuracy, that normalized global imbalance is an appropriate criterion for halting mesh refinement and that a specified level of accuracy is realized at lower computational effort using local refinement compared with a uniform finer mesh. © 1997 by John Wiley and Sons, Ltd.     

Turbulent channel flow with 2D wedges of random height on one wall

Direct numerical simulations (DNS) of a turbulent channel flow with 2D wedges of random height on the bottom wall have been performed. In addition, two other simulations have been carried out to assess the effect of the geometry on the overlying flow. In the first simulation, the four smallest elements were removed while in the other, a uniform distribution of wedges with the same area was used. Two Reynolds numbers were studied, Re_b=2500 and Re_b=5000 which correspond in case of smooth walls to Re_τ=180 and 300, respectively. Roughness on the wall induces separated regions, the reattachment occurring on the walls of the wedges or on the bottom wall. The pressure gradients on the walls increase the ejections and inrushes towards the wall. As a consequence the flow is more isotropic. The mechanism inducing an improved isotropy has been explained in term of the spectra and budgets of Reynolds stress. The comparison of the 3 surfaces has shown that near the wall, the uniformly distributed roughness represents only a poor approximation of the surface with wedges of random height. The Reynolds stresses, pressure distribution and spectra on the modified wall agree well with those on the random surface. Energy spectra show the pitch to height ratio of the largest elements to be the more appropriate geometrical parameter to describe the geometry.     

Scanning PIV investigation of the laminar separation bubble on a SD7003 airfoil

A laminar separation bubble occurs on the suction side of the SD7003 airfoil at an angle of attack α =  4–8° and a low Reynolds number less than 100,000, which brings about a significant adverse aerodynamic effect. The spatial and temporal structure of the laminar separation bubble was studied using the scanning PIV method at α =  4° and Re = 60,000 and 20,000. Of particular interest are the dynamic vortex behavior in transition process and the subsequent vortex evolution in the turbulent boundary layer. The flow was continuously sampled in a stack of parallel illuminated planes from two orthogonal views with a frequency of hundreds Hz, and PIV cross-correlation was performed to obtain the 2D velocity field in each plane. Results of both the single-sliced and the volumetric presentations of the laminar separation bubble reveal vortex shedding in transition near the reattachment region at Re = 60,000. In a relatively long distance vortices characterized by paired wall-normal vorticity packets retain their identities in the reattached turbulent boundary layer, though vortices interact through tearing, stretching and tilting. Compared with the restricted LSB at Re = 60,000, the flow at Re = 20,000 presents an earlier separation and a significantly increased reversed flow region followed by “huge” vortical structures.     

Some Reynolds number effects on two- and three-dimensional turbulent boundary layers

Experimental data for a two-dimensional (2-D) turbulent boundary layer (TBL) flow and a three-dimensional (3-D) pressure-driven TBL flow outside of a wing/body junction were obtained for an approach Reynolds number based on momentum thickness of Re _θ =23,200. The wing shape had a 3:2 elliptical nose, NACA 0020 profiled tail, and was mounted on a flat wall. Some Reynolds number effects are examined using fine spatial resolution (Δy ⁺=1.8) three-velocity-component laser-Doppler velocimeter measurements of mean velocities and Reynolds stresses at nine stations for Re _θ =23,200 and previously reported data for a much thinner boundary layer at Re _θ =5,940 for the same wing shape. In the 3-D boundary layers, while the stress profiles vary considerably along the flow due to deceleration, acceleration, and skewing, profiles of the parameter correlate well and over available Reynolds numbers. The measured static pressure variations on the flat wall are similar for the two Reynolds numbers, so the vorticity flux and the measured mean velocities scaled on wall variables agree closely near the wall. The stresses vary similarly for both cases, but with higher values in the outer region of the higher Re _θ case. The outer layer turbulence in the thicker high Reynolds number case behaves similarly to a rapid distortion of the flow, since stream-wise vortical effects from the wall have not diffused completely through the boundary layer at all measurement stations. Received: 9 June 2000/Accepted: 26 January 2001     

Numerical solutions of high-Re recirculating flows in vorticity-velocity form

A numerical method for computing high-Re laminar steady flows is presented. The incompressible Navier-Stokes equations are expressed in terms of vorticity-velocity variables, discretized in space by finite differences on a staggered grid and advanced in time by a scalar alternating direction implicit (ADI) procedure, which allows a fully vectorized computer code. The accuracy and efficiency of the present formulation are discussed in comparison with the standard ω-ψ and u, v, P forms. Numerical results are presented for two test cases: the driven cavity at Re up to 5000 and the backward-facing step at Re up to 800.     

Finite Element Solution of Laminar Flow through Square-Edged Orifice with a Variable Thickness

A numerical analysis of laminar flow through square-edged orifice has been studied for Reynolds number in the range 0 < Re₀ < 2000 with β values varying from 0.2 to 0.8 and with l* values varying from 1/16 to 1. It is shown thai the flow discharge coefficient gradually decreases when the orifice thickness/diameter ratio increases for a high porosity.     

Microscopic particle image velocimetry measurements of transition to turbulence in microscale capillaries

The character of transitional capillary flow is investigated using pressure-drop measurements and instantaneous velocity fields acquired by microscopic PIV in the streamwise–wall-normal plane of a 536 μm capillary over the Reynolds-number range 1,800 ≤ Re ≤ 3,400 in increments of 100. The pressure-drop measurements reveal a deviation from laminar behavior at Re = 1,900 with the differences between the measured and the predicted laminar-flow pressure drop increasing with increasing Re. These observations are consistent with the characteristics of the mean velocity profiles which begin to deviate from the parabolic laminar profile at Re = 1,900, interpreted as the onset of transition, by becoming increasingly flatter and fuller with increasing Re. A fully-turbulent state is attained at Re ≅ 3,400 where the mean velocity profile collapses onto the mean profile of fully-developed turbulent pipe flow from an existing direct numerical simulation at Re = 5,300. Examination of the instantaneous velocity fields acquired by micro-PIV in the range 1,900 ≤ Re < 3,400 reveal that transitional flows at the microscale are composed of a subset of velocity fields illustrating a purely laminar behavior and a subset of fields that capture significant departure from laminar behavior. The fraction of velocity fields displaying non-laminar behavior increases with increasing Re, consistent with past observations of a growing number of intermittent turbulent spots bounded by nominally laminar flow in macroscale pipe flow with increasing Re. Instantaneous velocity fields that are non-laminar in character consistently contain multiple spanwise vortices that appear to streamwise-align to form larger-scale interfaces that incline slightly away from the wall. The characteristics of these “trains” of vortices are reminiscent of the spatial features of hairpin-like vortices and hairpin vortex packets often observed in fully-turbulent wall-bounded flow at both the macro- and micro-scales. Finally, single-point statistics computed from the non-laminar subsets at each transitional Re, including root-mean-square velocities and the Reynolds shear stress, reveal a gradual and smooth maturation of the patches of disordered motion toward a fully-turbulent state with increasing Re.     

Aspect ratio effects on three-dimensional incompressible flow in a two-sided non-facing lid-driven parallelepiped cavity

A numerical study of the three-dimensional fluid flow has been carried out to determine the effects of the transverse aspect ratio, Ay, on the flow structure in two-sided non-facing lid-driven cavities. The flow is complex, unstable and can undergo bifurcation. The numerical method is based on the finite volume method and multigrid acceleration. Computations have been investigated for several Reynolds numbers and various aspect ratio values. At a fixed Reynolds number, Re=500, the three-dimensional flow characteristics are analyzed considering four transverse aspect ratios, Ay=1,0.75,0.5 and 0.25. It is observed that the transition to the unsteady regime follows the classical scheme of a Hopf bifurcation. An analysis of the flow evolution shows that, at Ay=0.75, the flow bifurcates to a periodic regime at (Re=600) with a frequency f=0.093 less than the predicted value in the cubical cavity. A correlation is established when Ay=0.5 and gives the critical Reynolds number value. At Ay=0.25, the periodic regime occurs at high Re value beyond 3500, after which the flow becomes chaotic. It is shown that, when increasing Ay over the unit, the flow in the cavity exhibits a complex behavior. The kinetic energy transmission from the driven walls to the cavity center is reduced at low Ay values.     

Numerical Simulation of the Flow Around an Infinitely Long Circular Cylinder in the Transition Regime

The development of three-dimensional structures and the succeeding transition to turbulence occurs in the wake of a circular cylinder at Reynolds numbers 190≤Re≤330. This regime is investigated numerically by means of a spectral element method. Earlier numerical works aimed mainly at reproducing characteristic wake patterns observed in experiments. Small sizes of computational domains and short integration times were chosen to save computational resources. Consequently, the quantitative results show a considerable scatter. Within this work, a step by step approach to highly accurate direct numerical simulations is described. Thorough studies of the effect of resolution and blockage are performed in the laminar, two-dimensional regime, resulting in Reynolds number relationships that exactly reproduce experimental data. Based on these results, a stability analysis is performed to obtain wavelengths that are unstable against spanwise perturbations and the critical Reynolds number for the onset of three-dimensionality. The most unstable wavelengths of the “mode A” and “mode B” instabilities and its multiples are used as periodicity length for direct numerical simulations. Effects of integration time, resolution in streamwise as well as spanwise directions, and periodicity length on the flow quantities are studied. Numerically obtained Reynolds number relationships of Strouhal number and base-pressure coefficients that fit accurately within measured results are given for the first time. Curves for drag and lift coefficients are provided and compared with previous numerical studies. Furthermore, physical interpretations of the wake transition are discussed. Since the separation of physical features and effects of experimental arrangements are frequently an open question, our numerical results are able to supply a contribution to the understanding of the physics of cylinder flow. Received 12 September 2000 and accepted 26 June 2001     

NUMERICAL MODELLING AND PARTICLE IMAGE VELOCIMETRY MEASUREMENT OF THE LAMINAR FLOW FIELD INDUCED BY AN ENCLOSED ROTATING DISC

The fluid flow field within an enclosed cylindrical chamber with a rotating flat disc was calculated using a finite volume computational fluid dynamics (CFD) model and compared with particle image velocimetry (PIV) measurements. Two particular laminar cases near the Transitional flow regime were investigated: Reynolds number Re=2.5×1 ⁴, chamber aspect ratio G (h/R_d)=0.2 and Re=4.2×10⁴, G (h/R_d)=0.217. This enabled direct comparison with the numerical and experimental results reported by other researchers. The computational details and some major factors that affect the computed accuracy and convergence speed are also discussed in detail. PIV results containing some 4300 velocity vector points in each of seven planes for each case were obtained from the flow field parallel to the rotating disc. It was found that PIV results could be obtained in planes within the boundary layers as well as the core flow by careful use of a thin laser illumination sheet and correct choice of laser pulse separation. There was close agreement between numerical results, the present PIV measurements and other reported experimental and numerical results.     

Numerical simulation of forced convection over a periodic series of rectangular cavities at low Prandtl number

Convective heat transfer in laminar conditions is studied numerically for a Prandtl number Pr = 0.025, representative of liquid lead–bismuth eutectic (LBE). The geometry investigated is a channel with a periodic series of shallow cavities. Finite-volume simulations are carried out on structured orthogonal curvilinear grids, for ten values of the Reynolds number based on the hydraulic diameter between Re_m = 24.9 and Re_m = 2260. Flow separation and reattachment are observed also at very low Reynolds numbers and wall friction is found to be remarkably unequal at the two walls. In almost all cases investigated, heat transfer rates are smaller than the corresponding flat channel values. Low-Prandtl number heat transfer rates, investigated by comparison with Pr = 0.71 results, are large only for uniform wall temperature and very low Re. Influence of flow separation on local heat transfer rates is discussed, together with the effect of different thermal boundary conditions. Dependency of heat transfer performance on the cavity geometry is also considered.     

Simulations of Turbulent Flow in a Plane Asymmetric Diffuser

Large-eddy simulations (LES) of a planar, asymmetric diffuser flow have been performed. The diverging angle of the inclined wall of the diffuser is chosen as 8.5°, a case for which recent experimental data are available. Reasonable agreement between the LES and the experiments is obtained. The numerical method is further validated for diffuser flow with the diffuser wall inclined at a diverging angle of 10°, which has served as a test case for a number of experimental as well as numerical studies in the literature (LES, RANS). For the present results, the subgrid-scale stresses have been closed using the dynamic Smagorinsky model. A resolution study has been performed, highlighting the disparity of the relevant temporal and spatial scales and thus the sensitivity of the simulation results to the specific numerical grids used. The effect of different Reynolds numbers of the inflowing, fully turbulent channel flow has been studied, in particular, Re _b  = 4,500, Re _b  = 9,000 and Re _b  = 20,000 with Re _b being the Reynolds number based on the bulk velocity and channel half width. The results consistently show that by increasing the Reynolds number a clear trend towards a larger separated region is evident; at least for the studied, comparably low Reynolds-number regime. It is further shown that the small separated region occurring at the diffuser throat shows the opposite behaviour as the main separation region, i.e. the flow is separating less with higher Re _b . Moreover, the influence of the Reynolds number on the internal layer occurring at the non-inclined wall described in a recent study has also been assessed. It can be concluded that this region close to the upper, straight wall, is more distinct for larger Re _b . Additionally, the influence of temporal correlations arising from the commonly used periodic turbulent channel flow as inflow condition (similar to a precursor simulation) for the diffuser is assessed.     

Optimized incompressible smoothed particle hydrodynamics methods and validations

The solution of the Poisson's equation used by the incompressible smoothed particle hydrodynamics (ISPH) methods for estimating the pressure field is expensive in CPU time. The CPU time, consumed by the inversion of the operator ∇(1/ρ∇) and the estimation of the right hand side of the Poisson's equation, increases with the number N of particles used in a purely Lagrangian framework. In this work, this default of ISPH methods is overcome by solving the Poisson's equation on a Cartesian grid. This SPH-mesh coupling is equivalent to the particle in cell method. In a first step, in order to analyze its efficiency, the optimized version of two ISPH methods (divergence free and density invariant) is compared with the standard weakly compressible SPH method through two benchmarks of incompressible bidimensional flows characterized by the Reynolds number Re, Lamb-Oseen vortex (10 ≤Re≤ 100) and lid-driven cavity flow (100 ≤Re≤ 1000). In a second step, the numerical results obtained by the three SPH methods are compared to laboratory experimental data of a dam break flow in order to show the performance of the SPH-mesh coupling in a practical and complex flow problem. As in the configuration of the experimental setup, the numerical results are obtained for a Reynolds number Re = 3.8 10⁶.     

Turbulent flow in a circular separationless diffuser at Reynolds numbers smaller than 2000

The results of experimental and numerical investigation of flow in a circular conical diffuser with a small conicity angle ensuring separationless flow are presented. The measurements are carried out in an air flow with the Reynolds number Re₂ in the diffuser exit section ranging from 600 to 3000. A considerable effect of the channel expansion on the flow pattern is found to exist. It is shown that, as distinct from a tube, in which only laminar flow can be realized as steady for Re < 2000, in the exit section of a diffuser with the generator slope of 0.3° and a length equal to 70 entry diameters a developed turbulent flow is formed for Re₂ > 1000. For Re₂ > 1300 this flow is steady, that is, almost independent of the turbulence level at the entry, and is determined by the Reynolds number Re₂ in the exit section. For Re₂ ≈ 1000 the turbulent flow continuously goes over into a laminar flow. The flow parameters measured at the diffuser exit correspond to calculations in accordance with the threeequation turbulence model.     

Numerical solutions of pulsating flow and heat transfer characteristics in a channel with a backward-facing step

The incompressible laminar flow of air and heat transfer in a channel with a backward-facing step is studied for steady cases and for pulsatile inlet conditions. For steady flows the influence of the inlet velocity profile, the height of the step and the Reynolds number on the reattachment length is investigated. A parabolic entrance profile was used for pulsatile flow. It was found with amplitude of oscillation of one by Re=100 that the primary vortex breakdown through one pulsatile cycle. The wall shear rate in the separation zone varied markedly with pulsatile flows and the wall heat transfer remained relatively constant. The time-average pulsatile heat transfer at the walls was greater as with steady flow with the same mean Reynolds number.     

Direct numerical simulation of turbulent flow and heat transfer in a square duct with natural convection

In this paper, a direct numerical simulation of a fully developed turbulent flow and heat transfer are studied in a square duct with an imposed temperature difference between the vertical walls and the perfectly insulated horizontal walls. The natural convection is considered on the cross section in the duct. The numerical scheme employs a time-splitting method to integrate the three dimensional incompressible Navier-Stokes equation. The unsteady flow field was simulated at a Reynolds number of 400 based on the Mean friction velocity and the hydraulic diameter (Re _m = 6200), while the Prandtl number (Pr) is assumed 0.71. Four different Grashof numbers (Gr = 10⁴, 10⁵, 10⁶ and 10⁷) are considered. The results show that the secondary flow and turbulent characteristics are not affected obviously at lower Grashof number (Gr ≤ 10⁵) cases, while for the higher Grashof number cases, natural convection has an important effect, but the mean flow and mean temperature at the cross section are also affected strongly by Reynolds stresses. Compared with the laminar heat transfer at the same Grashof number, the intensity of the combined heat transfer is somewhat decreased.