[en] Highlights: • Direct numerical simulation of the wake of a flat plate with a circular trailing edge and turbulent separating boundary layers (4 cases). • Understand the effect of increasing the ratio of the boundary layer momentum thickness to the thickness of the plate on shedding and other wake characteristics. • Increasing θ/D substantially resulted in greatly decreased shedding strength (shear-layer roll-up curtailed, shed vortices form within shear layer). • Shedding even seems intermittent by some measures (continuity of shedding demonstrated using a sensitive measure). • Shedding frequency varies noticeably (shown in the article to be the effect of high- and low-speed streaks in the upstream turbulent boundary layer). - Abstract: The near and very near wake of a flat plate with a circular trailing edge is investigated with data from direct numerical simulations. Computations were performed for four different Reynolds numbers based on plate thickness (D) and at constant plate length. The value of θ/D varies by a factor of approximately 20 in the computations (θ being the boundary layer momentum thickness at the trailing edge). The separating boundary layers are turbulent in all the cases. One objective of the study is to understand the changes in wake characteristics with changing θ/D (as obtained by decreasing D). Vortex shedding is vigorous in the low θ/D cases with a substantial decrease in shedding intensity in the largest θ/D case (for all practical purposes shedding becomes almost intermittent). Other characteristics that are significantly altered with increasing θ/D are the roll-up of the detached shear layers and the magnitude of fluctuations in shedding period. These effects are explored in depth. The effects of changing θ/D on the distributions of the time-averaged, near-wake velocity statistics are discussed.

[en] Highlights: • The numerical and the experimental results show an overall agreement. • The V-rib and the V-rib-impingement configurations increase the Nusselt number. • The V-rib-target configuration is effective due to the target surface enlargement. • The rib induced additional pressure loss is negligible. • The effect of V-ribs on the impingement plate is very promising. - Abstract: The secondary vortex structure of an impingement jet system is enhanced by V-ribs on both the impingement and target plates. Numerical and experimental investigations are conducted to study the flow field and heat transfer resulting from V-rib turbulators in an impingement cooling configuration. Three different cases are tested: V-ribs on both the impingement and target plates (V-rib), V-ribs only on the impingement plate (V-rib-impingement) and V-ribs only on the target plate (V-rib-target). The experiment is carried out on a 9 by 9 inline impingement array test facility. For the transient measurements, narrow band thermochromic liquid crystals (TLC) and thermocouples are applied to obtain the local heat transfer distribution. Pressure taps are used to measure the pressure loss. The numerical simulation is carried out with ANSYS CFX 14, using a steady state Reynolds-Averaged Navier-Stokes (RANS) approach and the Shear Stress Transport (SST) turbulence model. All studies are done for a Reynolds number range of 15,000 to 35,000. There is a good overall agreement between the experimental and numerical results for the cases studied. The detailed flow field from the numerical simulation is used to understand and complement the phenomena observed in the experiment. The evaluation of the flow field confirms that the V-ribs enhance the secondary flow structure in the impingement system and induce a positive heat flux ratio compared to the baseline case. Both experimental and numerical results show a Nusselt number increase for the V-rib-impingement and V-rib configuration, with a highest Nusselt number ratio of 1.16. Notice that the experiment cannot take the rib part into account due to the invalid 1D semi-infinite wall assumption there, while the CFD simulation allows for the consideration of heat transfer on the rib surface and thus complements the heat flux data on the target plate. Depending on the configuration, the CFD simulation shows a heat flux ratio of 1.06–1.34. The pressure loss of the system is comparable to the case with a smooth impingement plate and a smooth target plate.

[en] Highlights: • The paper proposes a novel semi-analytical model for the prediction of the onset of homogeneous condensation (i.e. Wilson point) in steam flows. • A new dimensionless number, i.e., the Wilson number Wi, is introduced to characterize the delay of condensation. • The analysis is based on both experimental data and numerical simulations. • The model is validated by means of four test cases. The deviation from the experimental measurements is in the range 1–10%, and lower than that of other more complex models in literature. • The model is applicable to the design of condensation-free supersonic nozzles. - Abstract: An investigation on non-equilibrium condensing steam flow is conducted to attain a semi-analytical model for the prediction of the Wilson point up to the critical point. The database for the analysis includes experimental observations in various nozzles and conditions (ranging from 20 to 150 bar) taken from the literature as well as additional data at lower and higher reduced pressures, generated by means of a calibrated quasi-1D model based on the method of moments. The simplified model is based on a reformulation of the Wilson point in terms of activation time, defined as the temporal interval between the instant at which the flow is at saturation conditions and the inception stable of condensation. This allows to incorporate the dependency of the Wilson point on the cooling rate and dew-point temperature, which are found the key parameters affecting the delay of condensation. The accuracy of the approach is proved by predicting the degree of subcooling on four different test cases, with deviations against experiments in the range of $1-10\%$. As demonstrated, the same approach can be exploited to design nozzles free of condensation.

[en] Highlights: • Flow characteristics of interior subchannels in a 61-pin hexagonal fuel bundle. • PTV and PIV optical-based measurements with matched-index-of-refraction. • Axial and transverse components of ensemble-averaged and RMS fluctuating velocity. • Compared two-point spatial velocity-velocity cross-correlations with prior work. • Discussion of integral length scales. - Abstract: This study summarizes the experimental effort to characterize the flow fields of various interior subchannels in a 61-pin wire-wrapped hexagonal fuel bundle prototypical for a sodium fast reactor. The objective was to generate high spatiotemporal velocity field data for computational fluid dynamics turbulence model validation. The experimental facility employed the matched-index-of-refraction and modern laser-based optical measurement techniques. It is the largest transparent hexagonal test fuel assembly. Measurements were performed in two planes parallel to the axial flow, Interior-1 and Center-2. The Interior-1 location captured fluid interactions in four narrower subchannels formed by the exterior row of pins near the hexagonal duct wall. The Center-2 location captured fluid interactions in two wider subchannels spanning from the center pin out to the hexagonal duct wall. All measurements have been performed at a Reynolds number of 19,000. Results include discussion about statistical convergence of the dataset, along with flow statistics such as ensemble-averaged velocity, root-mean-square fluctuating velocity, Reynolds stress, and integral length scales.

[en] Highlights: • We investigate heat transfer in laminar and turbulent pipe flows with different volume fraction of finite-size spherical particles. • We show that a considerable heat transfer enhancement (up to 330%) can be achieved in the laminar regime by increasing the volume fraction of spherical particles. • The heat transfer is observed to increase significantly in the laminar regime as the pipe to particle diameter ratio decreases for the parameter range considered here. • In the turbulent regime, only a transient increase in the heat transfer is observed before the process decelerates in time below the values in single-phase flows. • A heat transfer enhancement, measured with respect to the single phase flow, is only achieved at volume fractions as low as 5% in a turbulent flow. - Abstract: Controlling heat and mass transfer in particulate suspensions has many applications in fuel combustion, food industry, pollution control and life science. We perform direct numerical simulations (DNS) to study the heat transfer within a suspension of neutrally buoyant, finite-size spherical particles in laminar and turbulent pipe flows, using the immersed boundary method (IBM) to account for the solid fluid interactions and a volume of fluid (VoF) method to resolve the temperature equation both inside and outside the particles. Particle volume fractions up to 40% are simulated for different pipe to particle diameter ratios. We show that a considerable heat transfer enhancement (up to 330%) can be achieved in the laminar regime by adding spherical particles. The heat transfer is observed to increase significantly as the pipe to particle diameter ratio decreases for the parameter range considered here. Larger particles are found to have a greater impact on the heat transfer enhancement than on the wall-drag increase. In the turbulent regime, however, only a transient increase in the heat transfer is observed and the process decelerates in time below the values in single-phase flows as high volume fractions of particles laminarize the core region of the pipe. A heat transfer enhancement, measured with respect to the single phase flow, is only achieved at volume fractions as low as 5% in a turbulent flow.

[en] Highlights: • A new method for measuring particle concentration in turbulent flows is presented. • Quantitative, 3D data are obtained without optical access using a standard MRI. • Measurements in a homogeneous, turbulent channel flow agree well with theory. • Transport of a particle streak in a ribbed square channel flow is studied. - Abstract: A novel method, denoted MRP (short for Magnetic Resonance Particle concentration), was developed to obtain 3D volume fraction measurements for a dispersed particulate phase in turbulent water flows using Magnetic Resonance Imaging (MRI). MRI images taken near a single stainless steel particle suspended in agarose gel showed good agreement with the analytical solution for the disturbance to a uniform magnetic field induced by an immersed sphere. For a random distribution of particles, a linear relationship between the MRI signal decay rate ($R_2^{*}$) and particle volume fraction (${\varphi }_{\text{v}}$) has previously been predicted in the MRI literature. This relationship was investigated for various types of particles suspended in agarose gel vials. Good agreement with theory was observed for particles with a high magnetic susceptibility difference from water. $R_2^{*}$ was also measured in a square channel flow containing a uniform distribution of titanium particles at two fully turbulent Reynolds numbers. Experimental results again agreed well with theory in the majority of the channel for both Reynolds numbers studied. Data from this flow were used to examine the expected SNR and dynamic range for MRP in future experiments. Some discrepancy was observed near the entry region of the channel, with possible explanations including inflowing fluid and large-scale flow structure effects behind the channel’s mixing pin array. Finally, the new method was used to measure the 3D concentration distribution for a streak of titanium particles injected into a turbulent square channel flow with angled ribs. The transport of the streak was analyzed quantitatively, and a minor asymmetry in the channel geometry was shown to have important implications for the mean transport of the particle streak.

[en] Highlights: • Increasing slip effect increase Nusselt number in Br > 0. • Nusselt number decrease by increasing Brinkman number in Br > 0. • For the cooling case fluid starts to warm up in Br > Br₁. • Nusselt curve show a singularity in Br₂. • Br₁ and Br₂ increase by Increasing slip effect. - Abstract: Forced convective heat transfer in pipes is investigated for viscoelastic fluids obeying the Giesekus constitutive equation including effect of slip condition by an approximated analytical method. The slip equation at wall is considered as a nonlinear Navier model with non-zero slip critical shear stress. The problem under consideration is steady, laminar and fully developed. Thermal boundary conditions are assumed peripherally and axially constant heat flux at wall. The fluid heating and cooling cases are considered for analysis. Dimensionless temperature profiles and Nusselt number are obtained by solving governing equations and the effects of slip parameters, viscous dissipation and fluid elasticity are discussed. Results show that Nusselt number increases by increasing slip effect but decreases by increasing Brinkman number for the case of fluid heating. However, for the cooling case, the heat generated by viscous dissipation can overcome the effect of wall cooling at first critical Brinkman number and fluid starts to warm up. Also the Nusselt curve shows a singularity in a second critical Brinkman number.

[en] Highlights: • POD analysis carried out on velocity field data obtained in a cantilevered circular cylinder. • A new low frequency instability is identified for cantilevered cylinder flows of moderate aspect ratio. • The low frequency instability comprises a significant portion of the fluctuating energy in the wake. • The low frequency instability persists for laminar and turbulent vortex shedding conditions. • The instability is not observed if the incoming boundary layer is tripped, i.e, turbulent. - Abstract: Near-wake characteristics of a low aspect ratio ($h/d=4$) cantilevered circular cylinder protruding a thin laminar boundary layer were investigated both numerically ($Re=300$) and experimentally ($Re=10,400$). Despite the substantial differences in the investigated Re, the wake dynamics show striking similarities and appear governed by similar instability mechanisms: (i) a Kármán-like vortex shedding instability, and (ii) a low-frequency instability related to the flow over the free end and near the cylinder-wall junction. Attention is drawn to the low-frequency instability, which comprises a significant portion of the kinetic energy content in the wake, and has not been reported in previous experimental or numerical investigations. It appears to be characteristic of intermediate aspect ratio cantilevered circular geometries and the boundary layer state, since the phenomenon is not observed for turbulent boundary layers of similar thickness.

[en] Highlights: • At reattachment location , intense increase in crosswise velocity, normal stresses and higher mixing lead to a peak in heat transfer. • The separation/reattachment location depends upon specific stage of vortical structure and is a function of surface spacing. • Location of maximum value of mean cross-wise velocity and normal stress located at intersection of inner -outer shear layers. • The outer scaling is able to bring out the self-similar profile of mean velocity and normal stresses. • The results show that the outer scales are not suitable to scale the data in the inner layer. - Abstract: The flow field of an impinging wall jet created by the impingement of a turbulent axisymmetric jet normal to a flat surface was characterized by the particle image velocimetry technique. Experimental data is analyzed to explore two basic features of the impinging jet: first, to bring out unexplored aspects which are responsible for secondary peak in heat transfer distribution and to understand the reason for discrepancies in the existing observations about the peaks in heat transfer. Second, to analyze the self-similarity of radial wall jet based upon outer scaling. Measurements of the cross-wise mean velocity and turbulence statistics were initially used to explain the dynamics of secondary peak. Our results show that flow separation/reattachment occurs along the surface. At the reattachment location an intense increase in crosswise velocity, normal stresses and higher mixing are evident, which would lead to a peak in heat transfer. The separation/reattachment location is further found to depend upon the specific stage of vortical structure and is a function of surface spacing. The location of maximum value of mean cross-wise velocity and normal stress is located at the intersection of inner and outer shear layers. While the maximum Reynolds shear stress location is shifted to the outer shear layer and is located between the location of maximum velocity and jet half-width. The impinging jet exhibits a self-similar behavior as evident by the collapse of mean velocity and turbulent stress profiles when scaled with appropriate parameters. The outer scaling is able to bring out the self-similar profile of mean velocity and normal stresses. However, the shear stress profile does not show the self-similar behavior by the use of outer scaling. Data in the inner shear layer show small scatter compared to the outer shear layer especially close to the surface. The results show that the outer scales are not suitable to scale the data in the inner layer. It is also observed that the presence of vortical structure in wall jet delays attainment of self-similarity and the location beyond which self-similarity is observed is a function of surface spacing. These results aid in interpretation of heat transfer behavior from a flat surface and provide comprehensive benchmark data for theoretical modeling of the flow.

[en] Highlights: • Confined flashback limits are reproduced by LES with detailed chemical kinetics • The occurrence of boundary layer separation is independent of flame propagation • During flashback, the size of the separation regions exceeds the quenching distance - Abstract: Flame flashback from the combustion chamber into the upstream burner mixing section is an inherent threat associated with premixed combustion. Especially for hydrogen-containing fuels, boundary layer flashback (BLF) can occur and damage the combustor. It is thus important to develop methods for the accurate prediction of the onset of BLF. In this study, the ability of large eddy simulations (LES) to reproduce experimentally determined flashback limits of flames confined in a channel is investigated. First, inert LES of turbulent channel flow are conducted. Subgrid turbulence is modelled via the Smagorinsky turbulence model. The unsteady velocity profiles obtained from the inert simulations are then used as inlet boundary conditions for the reactive LES. The combustion process is modelled via finite rate chemistry and detailed chemical kinetics. By directly calculating the species reaction rates from filtered species and temperature distributions, turbulence-chemistry interaction of the subgrid scales is neglected. Differential diffusion and Soret diffusion are included in the species transport equations. It is shown that LES with the Smagorinsky turbulence model is capable of reproducing the turbulence characteristics of the channel flow. With the correct unsteady inlet velocity profile the chosen combustion model is capable of accurately predicting the onset of confined BLF. It can thus be concluded that the chosen combustion model reproduces all physical effects relevant to BLF. Furthermore, it is shown that BLF is initiated when the size of local flow separation regions clearly exceeds the quenching distance of the flame.