Search

Abstract: Micromixing of viscous systems in a stirred-tank reactor with Rushton turbine is investigated numerically. It is characterized by the product selectivity of parallel competitive reactions. Flow fields inside the reactor are determined by the Reynolds Stress Model (RSM). A computational fluid dynamics (CFD) method combining the standard E model and Finite-Rate/Eddy-Dissipation (FR/ED) model is implemented and is validated using experimental data in the literature. The simulations show that a higher agitation speed, a lower fluid viscosity and/or a feeding location closer to the discharge area of the impeller favor micromixing and the reaction rate. The trajectory of the reaction plume also influences the micromixing performance. The value of the FR/ED model parameter determined by lab-scale experiments decreases with increasing fluid viscosity. However, it is little affected by the agitation speed. These results provide useful guidelines for the scale-up of industrial reactors with complex chemical reactions. [Copyright &y& Elsevier]

Abstract: Large Eddy Simulations (LESs) were performed for a confined rectangular liquid jet with a co-flow and compared in detail with particle image velocimetry (PIV) measurements. A finite-volume CFD library, OpenFOAM, was used to discretize and solve the filtered Navier–Stokes equation. The effects of grid resolution, numerical schemes and subgrid models on the LES results were investigated. The second and fourth order schemes gave similar performance, while the fourth order scheme costs much more computationally. Subgrid model comparison showed that the locally dynamic procedure is necessary for complex flow simulation. Model validation was performed by comparing LES data for the point-wise velocity statistics such as the mean and the root-mean-square velocity, shear stress, correlation coefficient, velocity skewness and flatness with the PIV data. In addition, LES data for the two-point spatial correlations of velocity fluctuations that provide structural information were computed and compared with PIV data. Good agreement was observed leading to the conclusion that the LES velocity field accurately captures the important characteristics of all the turbulent length scales present in the flow, from the fully resolved energy-containing eddies to the subgrid-scale dissipative eddies. [Copyright &y& Elsevier]

Abstract: This paper proposes that a diesel oxidation catalyst (DOC) operating within the fully turbulent flow regime is an efficient means of reducing the contents of particulate matter in the exhaust gases. The suggested mode of operation is in contrast to the fact that the DOCs are typically operated within the laminar flow regime. In the paper, the particle trapping efficiency and pollutant conversion in turbulent ceramic DOCs are calculated using both mass-transfer correlations available in the literature and computational fluid dynamics (CFD). It is shown that a turbulent DOC substantially increases the removal of small particulates from the exhaust gases. This indicates the potential of the aftertreatment system to comply with the forthcoming number-based emission legislations on particulate matter. In addition, the turbulent DOC can be used to optimize the overall performance of a combined system consisting of a DOC and a diesel particulate filter. [Copyright &y& Elsevier]

Abstract: The bubble column reactors are usually operated in a heterogeneous regime where the liquid phase turbulence is generated by the bubble motion and the velocity gradients in the mean motion. The turbulent flow comprises of fluid elements moving in a random fashion with different sizes and energies, called ‘flow structures’. Both the large and small scale flow structures within a reactor play an important role in governing the local momentum, heat and mass transfer. The current work is focused on the estimation of the time averaged flow pattern and flow structures. The experimental data has been collected using miniature pressure sensors, PIV+shadowgraphy and LDA. The data was subjected/analyzed to/with multipoint linear stochastic estimation (MLSE), wavelet transforms, image processing and eddy isolation (EIM) to identify the flow structures. Two bubble columns have been used: a narrow rectangular (2D) column and a cylindrical (3D) column. Wavelet transforms (WT) were applied to isolate individual structures from PIV data to get their shape, size and energy in the 2D column. MLSE has been used to obtain the velocity profiles from pressure fluctuation signals. This data, augmented by PIV and LDA data, is subjected to WT and EIM to get the eddy age and its energy distribution. The data of the eddy shape size and energy was used to predict the mass transfer coefficient in the cylindrical bubble column as a test case. Overall, in this work we present a methodology to utilize the experimental data to get a better insight of the dynamics of flow structures, and propose a path forward for the deeper understanding of transport phenomena in bubble columns. [Copyright &y& Elsevier]

Abstract: A combined particle image velocimetry (PIV) and planar laser-induced fluorescence (PLIF) system was employed to investigate the turbulent mixing in a confined liquid-phase rectangular-wake flow with a Reynolds number based on hydraulic diameter of 37,500 and a Schmidt number of 1250. The simultaneous velocity and concentration field data were analyzed for flow statistics such as turbulent fluxes, turbulent viscosity and diffusivity, and turbulent Schimdt number. The streamwise and transverse turbulent fluxes were found to be of the same magnitude. The turbulent flux vector was not aligned with the mean concentration gradient. The turbulent Schimdt number was about 0.8. The spatial correlations of turbulent fluxes and concentration fluctuations were evaluated. In the correlation field, there were a positive and a negative vertically oriented elliptical correlation region, which were symmetric around the basis point. The correlation region was a horizontally oriented ellipse with negative values of the correlation coefficient. The correlation field of was also an ellipse with a horizontal major axis. The behavior of large-scale structures in both the velocity and concentration field was studied using linear stochastic estimation with a defined event of concentration fluctuation. Vortex streets were observed in the estimated velocity fields. The streamwise growth of the structure size increased linearly initially but then grew more slowly. [Copyright &y& Elsevier]

Abstract: A Lagrangian marker particle (LMP) method is applied to measure the toroidal large-scale eddies (LSEs) and their enveloping stagnation surfaces in a 280l bottom-sweeping model crystallizer. The trajectories of a 0.4cm diameter LMP show that these stagnation surfaces inhibit transport. Analysis shows that the velocity component normal to stagnation surfaces vanish. Therefore, stagnation surfaces act as a semi-permeable barriers to particle transport. Microconductivity measurements show that the stagnation surfaces are leaky at the molecular scale. Thus particle transport through stagnation surfaces is size-dependent. The LMP measurements reveal the structure of the LSEs. This consists of (1) an upward-swirling flow adjacent to the tank perimeter extending from the bottom to the top of the tank, (2) a central, quiescent zone, and (3) a downward return flow between (1) and (2) through a system of nested, smaller diameter, secondary toroidal flows concentric with the impeller axis. A cylindrical stagnation surface surrounds the central quiescent zone. These results are corroborated by measurements of inhomogeneous concentration profiles in an industrial scale 2000l batch crystallizer. This leads to an understanding of the effects of LSEs on silver halide microcrystal particle size distribution in the industrial scale crystallizer. [Copyright &y& Elsevier]

Abstract: Turbulent fluid flow and related solid particle behaviour in the direct vicinity of the heat exchanging (HE) surface of a scraped heat exchanger crystallizer was studied. The liquid flow is visualized by dye injection and the particles are monitored directly for two types of commonly used scraper geometries. In conjunction with this experimental work, we performed direct numerical simulations of the two-phase (solid–liquid) flow system. Our main goal is the design of scraper geometries that enhance heat transfer by perturbing the thermal boundary layer, and effectively scrape off particles that nucleate, grow and adhere onto the HE surface. Also the turbulent flow generated by the moving scrapers should direct the particles into the bulk of the tank. The experiments and simulations show good qualitative resemblance which enables the design of scrapers based primarily on numerical predictions. [Copyright &y& Elsevier]

Abstract: Large regions of inhomogeneous mixing have been observed in industrial, bottom-sweeping impeller crystallizers. To investigate this phenomenon, we conducted experiments on a one-tenth volume model of this kind of mixing tank. Results are reported of Lagrangian marker particle (LMP) and microconductivity measurements using the model mixing tank with impeller tip Reynolds numbers of 25,000. Surprising structure is found in this high Reynolds number flow. Using the LMP trajectory data, we show the flow consists of a narrow region of rapidly moving, upward spiralling flow at the tank perimeter. This flow returns slowly through a vertical stack of tori and through a quiescent region centered on the impeller. These tori are concentric with the impeller and exist at loci of regions of shear found adjacent to the quiescent central region and the tank perimeter. Conditional analysis of the microconductivity signals reveals that large concentration fluctuations occur in the perimeter flow. In contrast, only small diffusive-like concentration fluctuations occur in the center of the tank. This segregation of regions of rapid transport in the perimeter flow from regions of micromixing in the quiescent region results in inhomogeneous mixing in the tank. The complexity of the flow is reflected in the large dynamical dimension () of the flow obtained from the calculation of the Kolmogorov entropy production rate. The return-time distribution was found to be composed of a superposition of two log-normal distributions. Period doubling phenomenon was also found in these distributions. [Copyright &y& Elsevier]

In order to represent radial dispersion phenomena in packed beds a three-dimensional network model is proposed based on characteristic structural properties of the packing. As shown by a detailed sequential experimental study, the model is able to predict radial Peclet numbers for the turbulent regime in good agreement with experimental data without adjusting any model parameter. [Copyright &y& Elsevier]

Abstract: The use of eccentric agitation, where the impeller is placed away from the centreline axis, was examined as a means of obtaining effective mixing in overhead-stirred unbaffled vessels used as high throughput experimentation (HTE) reactors. The performance was compared with conventional baffled and unbaffled configurations where the impeller was positioned at the vessel axis. Three vessels of diameters 35, 45 and 60mm were used which comprehensively covers the HTE scale of interest. Power input per unit volume was maintained at a constant 168Wm for all experiments and water was used as the working fluid. Gross flow circulation and mixing times were determined by use of particle image velocimetry and planar laser induced fluorescence, respectively. The spatial distribution of turbulent kinetic energy within the vessels was determined to be independent of scale. However, local values of the turbulent kinetic energy within the spatial distribution were found to be scale-dependent. Mixing time has been found to be a function of vessel diameter, with the smaller vessels shown to be more efficient mixers. Energy dissipation within the vessels was estimated using a sub-grid scale closure model. The Smagorinsky SGS closure model was selected in preference of traditional estimation models, however all the models used gave a poor prediction of energy dissipation for the flow conditions studied. [Copyright &y& Elsevier]