When a material spreads in turbulent flow, areas with high concentration are formed by the local turbulence. These high concentration islands can lead to heavy damage. In this study, the high concentration areas are extracted as pulses by a conditional sampling technique, and statistical analysis were performed in order to investigate the difference between the pulses and concentration plume. Assuming homogeneous flow, a series of images by Planar Laser Induced Fluorescence and Particle Image Velocimetry techniques are acquired at several downstream positions. For PLIF images, a conditional sampling technique is used to extract the high concentration areas. For simultaneous PIV and PLIF images, wavelet analysis is performed to probe into the relationship between high concentration pulses and velocity fluctuation. Turbulent diffusion coefficient of the high concentration areas is found to increase with the downstream distance whereas that of the concentration plume is constant, which implies that the diffusion process of the pulses differs from that of the plume. Besides by wavelet analysis result, it is indicated that concentration fluctuation in the pulses is negatively correlated with streamwise velocity fluctuation. It is concluded that high concentration pulses and plume have different generative mechanisms.

The present experiment focuses on the vorticity diffusion in a stronger wall jet managed by a three-dimensional flat plate wing in the outer layer. Measurement of the fluctuating velocities and vorticity correlation has been carried out with 4-wire vorticity probe. The turbulent vorticity diffusion due to the large scale eddies in the outer layer is quantitatively examined by using the 4-wire vorticity probe. Quantitative relationship between vortex structure and Reynolds shear stress is revealed by means of directly measured experimental evidence which explains vorticity diffusion process and influence of the manipulating wing. It is expected that the three-dimensional outer layer manipulator contributes to keep convex profile of the mean velocity, namely, suppression of the turbulent diffusion and entrainment.

Effects of system rotation are investigated on inhomogeneous flow where disturbances are transported from turbulent to non-turbulent flow field through advection and turbulent diffusion. With the body force on the fringe region, spectral method is used for inhomogeneous flow which is stirred at the bottom of the cuboid computational box. As a result, it is found that inertial waves with a constant helicity are transmitted in the direction perpendicular to the stirred surface, and parallel to rotational axis. In this study, the effects of system rotation on generation and propagation of wave are discussed.

The mixing of liquid mass caused by a spherical solid particle falling in a still liquid in a pipe was investigated by visualization and noninvasive concentration measurement using a photochromic dye. A spherical particle with diameter of 4.76 mm was dropped in a kerosene-paraffin mixture liquid with a photochromic dye. The photochromic dye was activated by an ultraviolet light and was subjected to the mixing by the particle wake. The falling velocity of particle was changed by using 8 different densities of particle. The effect of the particle Reynolds number on the mixing was investigated for the Reynolds number range from 10 to 2490. The effect of liquid viscosity on the mixing time was also investigated using two liquids having different viscosity. The visualized dye patterns indicated the mixing process depended strongly on the particle Reynolds number. For the Reynolds numbers higher than 300, the particle shed the vortices behind the particle and the dye was mixed isotropically by large-scale vortices. For the Reynolds numbers lower than 300, the dye was drawn straightly by a laminar wake of the particle. The concentration of the mixed dye was measured using the photochromic concentration measuring (PCM) technique to discuss the mass mixing quantitatively. The turbulent diffusion coefficient (TDC) was evaluated for the cases the dye was mixed by the vortices. It was found that the evaluated TDCs showed strong time-dependency, which was attributed to the change in scale and whirling velocity of wake vortices. The maximum TDC depended on the falling velocity regardless of the fluid viscosity. The mixing time depended strongly on the liquid viscosity. The mixing time of the TDC was suggested to be governed by the viscous decay time and expanding time of vortices in the pipe. The amount of dye drift was evaluated for the cases the particle wake was laminar. It was found that the dye drift increased sharply just after the particle passing and then saturated. The final dye drift increased gradually with increasing Reynolds number.

A numerical investigation of two-phase solid-liquid (slurry) flow in horizontal pipes has been carried out. Simulations of concentrated slurry flows in pipes 0.0515 m and 0.15 m in diameter were performed using the two-fluid approach implemented in the commercial CFD code, ANSYS CFX. Mixtures of monosized and bimodal particle sizes were tested. Several test cases were investigated to predict particle velocity- and concentration-distributions and frictional pressure gradients. The effects of turbulence model selection, dispersed phase wall boundary conditions, and interphase force terms on model performance were evaluated. The selection of turbulence model had a significant impact on the dispersed phase velocity and concentration distributions. Comparison of simulations with benchmark experimental data shows clearly that for the relatively small particle sizes (∼100 microns), poor solids concentration profile predictions are obtained if the turbulent dispersion force is not included. In general, very good agreement between numerical and experimental results was observed.

In this study, simultaneous measurements of velocity and concentration fields using the time-resolved particle image velocimetry (PIV) and planar laser induced fluorescence (PLIF) methods have conducted to investigate mixing characteristics in turbulent water flows driven by air bubbles in a cylindrical water tank. The flow rates of compressed air is changed from 1 to 5 L/min at 0.5 MPa and the corresponding range of bubble based Reynolds number is from 8,320 to 22,100. PLIF measurement results demonstrate that the mixing efficiency is enhanced with increase of gas flow rate. The sloshing motion of the free surface is also effective to the scalar mixing process since the vertical motion can be correlated with concentration fluctuation and increase turbulent dispersion process.

The interface between the turbulent and non turbulent regions in a two dimensional turbulent jet is investigated by the simultaneous measurement of the velocity and pressure. The measurement is performed by using a combined probe comprising an X-type hot-wire and a static pressure tube. The measurement data are analyzed by the conditional sampling technique and an ensemble averaged technique on the basis of the intermittency function for the turbulent/non turbulent decision. The experimental data at the cross-streamwise edge of the turbulent region show that there is a thin interfacial layer with a sharp jump of physical quantities (such as mean streamwise velocity) at the cross-streamwise edge of the turbulent region, and the thickness of the interfacial layer is 0.08 times the half-width of the cross-streamwise profile of the mean streamwise velocity. The turbulent diffusion term in the turbulent energy transport equation near the interfacial layer is examined. It is also found that the turbulent energy is transported from the inside of the interfacial layer to both the inner side (the side of the turbulent fluid) and the outer side (the side of the non turbulent fluid) by the diffusion term. Furthermore, the components of the diffusion term are separately estimated. It is found that the turbulent diffusion term shows the gain of the turbulent energy at the inner side of the interfacial layer, and the pressure diffusion term transports the turbulent energy to the non turbulent fluid. Moreover, small scale vortices are found in the interfacial layer. From these results, there is a possibility that the existence of the interfacial layer (existence of the vortices) contributes to the transport of the turbulent energy to the non turbulent fluid since the velocity and the pressure field that determine the pressure diffusion is greatly influenced by the existence of the interfacial layer. This hypothesis indicates that the outward propagation from a turbulent fluid can be attributed to the presence of vortices in the interfacial layer.

The turbulent mixing of liquid mass caused by an air bubble rising near a wall in a still liquid in a pipe is investigated experimentally using a photochromic dye. A part of the liquid is activated by UV light and subjected to the fluid motion caused by a zigzag rising bubble of which Reynolds number is 214. The visualized mixing patterns showed that the dye is mixed by vortex motions in the bubble wake that is similar to the case of a bubble rising in the center of the pipe. The concentration distributions were deduced from the dye images using Lambert-Beer’s law and the turbulent diffusion coefficient (TDC) was evaluated from the temporal changes in the mass dispersion. The TDCs showed that a near-wall bubble generates stronger mixing than for a bubble in the center of the pipe. This stronger mixing can be attributed to the large-scale vortices observed for a near-wall bubble, which remains active for a longer time due to the lack of oppositely rotating vortices and mixes more fluids.

The efficacy of pesticide sprays is strongly dependent on the accuracy of the droplet size spectrum. Over estimates of droplet size may result in excessive drift or ineffective doses, while under estimates of droplet size result in excessive releases. This situation is not only bad for the environment; it incurs large operating costs (spray usually accounts for 30% of total cost). This paper describes the study of droplet sprays commonly used in the agriculture and forestry management. It combines experimental wind tunnel testing and Computational Fluid Dynamic (CFD) methods to develop a fundamental understanding of droplet generation and dispersion in the wake of the atomizer spray system. The results will assist designers of spray technology and applicators in delivering pesticide to its target. The CFD models that are developed and calibrated will further allow the wind tunnel data to be generalized; thus, allowing less wind tunnel testing and eventually direct simulation of droplet dispersion in aircraft wakes. The CFD models are developed for the poly-dispersed sprays released from a Micronair AU4000 atomizer (a standard atomizer used for forest protection) at an airspeed of 67 m/s. Simulations are performed using a Lagrangian (droplet phase) - Eulerian (fluid phase) procedure and include droplet drag/body forces and turbulent dispersion of droplets. The Base-line Reynolds Stress Model (BSL RSM) turbulence model is used to compute turbulence levels in the air phase. The CFD simulations include the sprayer and a large portion of the wind tunnel geometry in order to facilitate in validation. The computational results are compared to full scale experimental measurements of pressure, gas phase velocity, droplet velocity, and droplet size spectra measured using Phase Doppler Interferometry (PDI) and Hotwire Anemometry. Measurements are available along radial lines at 0.5, 1, 2 and 4 m downstream of the atomizer.

A continuous random walk (CRW) turbulent diffusion model was developed for Lagrangian particles within flow fields simulated by hybrid RANS/LES methodologies. For RANS flow-fields, the conventional time-scale and length-scale constants were determined by the turbulence intensity and dissipation values computed by the single-phase solver with a k-ω (Menter SST) model and subsequent comparison with turbulent particle diffusion experimental results of Snyder & Lumley (1971). This allowed validation against data for four particle types ranging from hollow glass to copper shot in grid-generated turbulence. The stochastic diffusion model was then extended to utilize the Nichols-Nelson k-ω hybrid RANS-LES turbulence model in a more complex turbulent flow resulting from the unsteady, three dimensional wake of a cylinder at Mach number of 0.1 and Reynolds number (Re D ) of 800. The gas flow was computed with a 5 th -order upwind-biased scheme. Throughout the wake, the sub-grid random walk model yielded good predictions of particle diffusion as compared with DNS. Also, these results indicate that crossing trajectory effects and inertia-based drift corrections are critical to handling a variety of particle Stokes numbers as well as regions of non-homogeneous turbulence, even when most of the kinetic energy is captured with the resolved-scales of an LES approach.

Large eddy simulation (LES) using a mixed-time-scale (MTS) SGS model is applied to the intake flows in simplified internal combustion engine geometry. A modified colocated grid system is employed so as to perform calculations stably with a central difference scheme for convection terms. The results are compared with corresponding experimental data and the computational results obtained by using the low-Reynolds-number linear k-ε and cubic nonlinear k-ε -A 2 turbulence models. The LES results show the best agreement with experimental data in three computational cases, not only in the mean velocity profiles but also in the profiles of turbulent energy. These results suggest that LES using the MTS model is an effective method for accurately predicting the performance of a combustion engine involving turbulent diffusion of spray and combustion flame. In addition, it is clarified that the numerical viscosity from the QUICK scheme for the convection term has a great influence on the computational results and decreases the prediction accuracy of LES.

This paper describes an application of the partial chemical equilibrium method considered chemical kinetics in computational fluid dynamics (CFD). In this method, fuels and oxidants are mixed at a turbulent rate so that a mixture gas of fuel and oxygen is generated. Next, the mixture gas of fuel and oxygen is burnt by molecular diffusion thereby resulting in combustion gases. The turbulent mixture rate is estimated by the eddy dissipation model and the burning velocity is evaluated by the Arrhenius equation. Finally, the combustion products are calculated by the chemical equilibrium method by using the combustion gases. One of the advantages of this method is its ability to calculate the combustion products without using chemical equations. The chemical equilibrium method requires only thermo-chemical functions (specific heat, standard enthalpy, etc). This method can be applied to incinerators or some complex combustion instruments and it can predict the intermediate chemical species of dioxins, etc.

The purpose of the present communication is to summarize the state-of-the art knowledge of the existing relationships between Eulerian and Lagrangian statistics in the literature. It will focus on integral scales, especially the classical Eulerian scales (such as those classically obtained from a fixed hot-wire or LDA control volume), the interesting moving Eulerian time scale (that obtained by a sensor which would move with the mean velocity) and the Lagrangian integral scale, most important for Lagrangian turbulent diffusion and dispersion simulations.

A grid-free method for the LES of turbulent incompressible flow is presented. The computational engine is based on the Lagrangian Vortex Element Method to account for the dynamics of the resolved vorticity field and the Vorticity Redistribution Method (VRM) to account for turbulent diffusion. Turbulence is modeled here using the standard Smagorinsky subgrid-scale model but work is underway to implement a dynamic version. In this paper, the accuracy and convergence rate of VRM are presented for the case of diffusion with variable viscosity. Furthermore, the adaptivity as well as the robustness of the proposed method in capturing large-scale vortex structures typical of vortex dominated turbulent flow are demonstrated using preliminary results from the LES of the collision of a pair of coaxial vortex rings, perturbed initially by the wave-number of their most unstable mode.

When a particle moves in a fluid, the fluid is disturbed by the particle and the fluid mass is mixed. This phenomenon is commonly observed in particle-laden flows and dispersed bubble flows. This mass mixing can be composed of two mechanisms. One is the mass transfer by convective flows that are induced by the reaction of the particle drag and the other is the turbulent mass mixing in the turbulent wake of the particle. The effect of the former one can be evaluated using the previous studies on the particle drag. However, the effect of the turbulent mixing is little understood. The turbulent mixing caused by a particle wake is investigated by visualization and noninvasive concentration measurement using a photochromic dye. A sphere brass particle of 5mm in diameter is dropped in kerosene filled in a vertical pipe and the mixing of dye is visualized. The photochromic dye, which is activated by an ultraviolet light, keeps its color in a few minutes after the activation. A part of the fluid is activated without disturbances and is subjected to the mixing by the particle wake. The visualized dye patterns indicate the dye is mixed isotropically by large-scale vortex motions when the particle sheds the vortices. Furthermore, the photochromic concentration measuring (PCM) technique is developed to obtain the concentration of the mixed dye. This PCM technique is based on the Lambert-Beer’s law for light adsorption and provides the average dye concentration along the light path. The measured concentration distribution shows rather isotropic mixing in longitudinal direction. The turbulent diffusion coefficient (TDC) is calculated from the temporal changes in the measured concentration distributions. The evaluated TDC shows strong time-dependency, which is attributed to the change in scale and strength of wake vortices. The TDC is about 10 4 times larger than the molecular diffusion coefficient at its maximum. The effect of particle Reynolds number on the turbulent mixing is also investigated for the Reynolds number range from 263 to 3290. The observed mixing patterns show a drastic change at the critical Reynolds number on the vortex shedding from the particle. The Reynolds number dependency on the non-dimensional TDC and mixing time are discussed.

Two-fluid model simulations of a bubbly vertical jet are presented. The purpose of these simulations is to assess the modeling of turbulence dispersion and lift forces in a free shear flow. Although turbulence dispersion forces have previously been validated using simpler canonical flows and microscopic particles or bubbles, there was a need to asses the model performance for larger bubbles in more turbulent flows. This method, of validating two-fluid models in flows of increasing complexity has the advantage of excluding, or at least minimizing, the possibility of cancellation of errors when modeling several forces. In a companion paper (see Part-II), the present two-fluid model is extended to a boundary layer in which forces induced by the presence of a wall are important. The turbulent dispersion models used herein are based on the application of a kinetic transport equation, similar to Boltzmann’s equation, to obtain the turbulent diffusion force for the dispersed phase [1, 2]. They have already been constituted and validated for the case of particles in homogeneous turbulence and jets [3] and for microscopic bubbles in grid generated turbulence and mixing layers [4]. It was found that it is possible to simulate the experimental data in Ref. [5] (See Figures-1 to 4) for a bubbly jet with 1 mm diameter bubbles. Good agreement is obtained using the model of Brucato et al. [7] for the modulation of the drag force by the liquid phase turbulence and a constant lift coefficient, C L . However, little sensitivity is observed to the value of the lift coefficient in the range 0 < C L < 0.29.

Three-dimensional measurements of bubble size, and velocity are presented along with the associated statistics for dilute bubbly flow in a vertical pipe. These measurements were obtained through a combination of precise index-of-refraction matching and a stereoscopic high-speed imaging system. An image-processing algorithm has been developed and used to extract instantaneous bubble size, shape, velocity and trajectory information and statistics corrected for magnification errors (maximum of 2%). The experiments have been conducted at a Reynolds of 14,600 in vertical pipe-flow and a volumetric gas flow ratio of 0.29%. Bubble sizes ranged from sub millimeter to several millimeters. The results indicate that bubbles less than approximately 20 wall units with very low ratios of rise velocity-to-liquid-average-velocity occupy the full breadth of the test section, while bubbles in the range between 20 and 60 wall units with ratios of rise-velocity-to-liquid-average-velocity greater than 0.5 are efficiently trapped by the wall and are almost exclusively found in the inner wall region (y + < 30). Bubbles larger than 60 wall units with ratios of rise-velocity-to-liquid-average-velocity less than 0.5 are found throughout the pipe cross-section although with a strong preference to the wall neighborhood. This bubble behavior is put in perspective considering four mechanisms of bubble migration: (a) turbulent dispersion due to near-wall large scale structure, (b) inviscid transverse forces (Magnus), (c) viscous transverse forces (Saffman), and (d) unsteady transverse forces because of vortex shedding and bubble shape deformation.

Turbulent transport of Reynolds stress by triple moment of fluctuating velocity and pressure-velocity correlation are evaluated from a DNS for wake of a rectangular cylinder located in a uniform flow. The turbulent diffusion transport of turbulent kinetic energy hardly correlates with its gradient vector, indicating that the generally accepted gradient diffusion model is inadequate. It is inferred that the separate modeling of the individual Reynolds stress component provides a better possibility for modeling the turbulent transport in this flow, because of the strong departure from the equilibrium state of turbulence. The possibility of the direct modeling of the pressure gradient-velocity correlation is discussed.

A mathematical model to simulate boiling flows in industrial applications is presented. Following the Eulerian multifluid framework, separate sets of mass, momentum, and energy conservation equations are solved for liquid and vapor phases, respectively. The interactions between the phases are accounted for by including relevant mass, momentum, heat exchanges and turbulent dispersion effects. Velocity-pressure coupling is achieved through a multiphase version of the SIMPLE method and the standard k-ε turbulence model is employed. In order to validate and assess the accuracy of the boiling model, subcooled nucleate boiling flows in a vertical annular pipe are simulated in the steady-state mode. The computed axial velocities, volume fractions, temperature profiles are compared with available experimental data (Roy et al., ASME J. of Heat Transfer, Vol. 119, 1997). The result obtained by assuming a constant value for the bubble diameter shows a reasonable agreement, but several limitations are observed in the details. A more advanced mathematical model incorporating separate transport equations for the bubble number density and the interfacial area is suggested.

Time-dependent simulations of a particle-laden swirl flow in a pipe expansion based on the Euler-Lagrange approach are presented. Two equation and Reynolds Stress Models were used in the calculation of turbulent quantities in the continuous phase. Additional attention was payed to the influence of particle dispersion. The instantaneous fluid velocities seen by the particles was reconstructed by different dispersion models. To come to a time dependant solution for the Euler-Lagrange approach, a quasi-unsteady approach is taken. This results in a calculational scheme where one Eulerian time-step is divided in a number of Lagrangian steps. Particle source term are sampled which represent the influence of the disperse phase on the flow field. which call for additional coupling within one Eulerian time step. The effect of inter-particle collisions on the movement of the disperse phase is accounted for using a stochastic inter-particle collision model. Special interest of this study was the formation of dust ropes which are observed in such flows.