Laser Doppler velocimetry measurements and computational fluid dynamic (CFD) simulations of turbulent flows with free-surface vortex in an unbaffled dish-bottom stirred tank reactor agitated by a Rushton turbine are presented. Measurements of the three mean and fluctuating components of the velocity vector are made in order to characterise the flow field and to provide data for CFD model validation. An Eulerian-Eulerian multiphase flow model coupled with a volume-of-fluid method for capturing the gas-liquid interface is applied to determine the vortex shape and to compute the flow field. Turbulence is modelled using the standard k-e, shear-stress transport and the differential Reynolds-stress model with two variants of the pressure-strain correlation. The predicted mean flow field obtained using all four turbulence models are on the whole similar and generally in good agreement with measurements. However, the Reynolds-stress models provide somewhat better predictions of the mean axial velocity. The turbulent kinetic energy is well predicted in the flow below the impeller, near the bottom of the tank; whereas it is underpredicted in the region close to the impeller and near the wall by all turbulence models.

This is an empirical examination of the association between community cohesion and the pattern of e-mail among rural Canadians. Residents of 27 diverse Canadian rural communities were invited to complete an Internet survey that included questions about the frequency of e-mail and the Buckner Scale subdimension of psychological sense of community (PSOC). Controlling for demographic variables including age, residence, income, education, language, and gender, as well as computer skill, and as based on ordinary least squares regressions, PSOC predicted local, regional, and provincial e-mail use but not national or international communication. The effects of PSOC declined with distance. Results suggest that e-mail use could be increased by developing greater computer skills; it could also be increased by changing the attitudes that people have as measured in PSOC.

The pharmaceutical and fine chemicals industries are strongly concerned with the manufacture of high value-added speciality products, often in solid form. On-line measurement of solid particle size is vital for reliable control of product properties. The established techniques, such as laser diffraction or spectral extinction, require dilution of the process suspension when measuring from typical manufacturing streams because of their high concentration. Dilution to facilitate measurement can result in changes of both size and form of particles, especially during production processes such as crystallisation. In spectral extinction, the degree of light scattering and absorption by a suspension is measured. However, for concentrated suspensions the interpretation of light extinction measurements is difficult because of multiple scattering and inter-particle interaction effects and at higher concentrations extinction is essentially total so the technique can no longer be applied. At the same time, scattering by a dispersion also causes a change of phase which affects the real component of the suspension's effective refractive index which is a function of particle size and particle and dispersant refractive indices. In this work, a novel prototype instrument has been developed to measure particle size distribution in concentrated suspensions in-process by measuring suspension refractive index at incidence angles near the onset of total internal reflection. Using this technique, the light beam does not pass through the suspension being measured so suspension turbidity does not impair the measurement.

Measurements and numerical simulations of turbulent flows with free-surface vortex in an unbaffled reactor agitated by a cylindrical magnetic stirrer are presented. Measurements of the three mean and fluctuating components of the velocity vector are made using a laser Doppler velocimetry in order to characterise the flow field at different speeds of the stirrer. A homogeneous Eulerian–Eulerian multiphase flow model coupled with a volume-of-fluid method for interface capturing is applied to determine the vortex shape and to compute the turbulent flow field in the reactor. Turbulence is modelled using a second-moment differential Reynolds-stress transport (RST) model, but for some cases the k–ε/k–ω based shear-stress transport (SST) model is also used. The predictions obtained using the ANSYS CFX-5.7 computational fluid dynamics code are compared with the images of the vortex and the measured distributions of mean axial, radial and tangential velocities and turbulent kinetic energy. The predicted general shape of the liquid free-surface is in good agreement with measurements, but the vortex depth is underpredicted. The overall agreement between the measured and the predicted axial and tangential velocities obtained with the RST model is good. However, the radial velocity is significantly underpredicted. Predictions of the turbulent kinetic energy yield reasonably good agreement with measurements in the bulk flow region, but discrepancy exists near the reactor wall where this quantity is underpredicted. The SST model predictions are generally of the same quality as those of the RST model, with the latter model providing better predictions of the tangential velocity distribution.

A standard method to determine particle shape and size is by image analysis. This paper addresses microscopic image analysis (semi-automated) investigations of two different organic crystalline chemicals generated by batch cooling crystallisation. The results generated from microscopic image analysis were compared with data obtained by dynamic image analysis (automated) because very few contributions are available in file open literature. The chemical systems were polymorphic L-glutamic acid which crystallises into a (prismatic) or (needle) form and the non-polymorphic mono sodium glutamate which crystallises into needles. The images from these techniques were processed to generate information oil crystal shape and size. It has been observed that shape effects can distort the size obtained in size characterization studies. In this study, comparisons were made of processing time. number of crystals and accuracy between microscopic and dynamic analysis. For representative microscopic image analysis, 5000 crystals were analysed in an average of eight hours while several hundred thousand crystals were processed using dynamic image analysis within 15 minutes. Using the parameters D(10), D(50), D(90), span and aspect ratio for statistical comparison, it was found that the results obtained for D50 by the two techniques were comparable and in accordance with other measurements (laser diffraction spectroscopy and ultrasonic attenuation spectroscopy) even through these non-spherical particles had different orientations during measurement by the two methods. However, Substantial differences in span of the distribution and aspect ratio were returned by the two techniques.

One thousand nine hundred ninety-five individuals in 20 rural Canadian communities were measured on perceived social cohesion by the three Buckner scale subdimensions: psychological sense of community (PSOC), attraction, and neighboring. Number of household children, income over $20,000, age, birthplace in, and years lived in the community significantly positively influenced PSOC and Attraction. Number of household children (positive for income over $20,000; otherwise negative), income over $40,000, birthplace, and years in the community significantly influenced neighboring. Increased interaction generally increases individuals' social cohesion. As the only significant community variable was being on an island province, individual-oriented policies are recommended to increase cohesion.

Batch-cooling crystallization is an important industrial unit operation often carried out from aqueous solutions. Measurement of crystal size and shape plays a major role in crystallizer control in order to improve product quality. Certain crystals grown from solution may exhibit polymorphism, which can significantly affect product properties such as bioavailability as well as impinging on downstream operations such as filtration and drying L-Glutamate was considered as a model compound for this study because it crystallizes from aqueous solution into two polymorphic forms, alpha and beta, which are rhombic and acicular respectively. In this study, cooling rate and initial solution concentration were chosen as manipulated variables to control polymorph formation. Rapid cooling from low solution concentration favors the formation-of the a form, while slow cooling with high concentration favors the beta form. Oxalic acid dihydrate and copper sulfate pentahydrate, which crystallize into monoclinic and triclinic systems, respectively, are also used in this study. Crystal morphology interacts with other product quality measurements, particularly crystal size. Currently, acquiring crystal size along with shape measurement is not readily achieved. Image analysis measures shape with size distribution but only for very small samples, and it is not currently practical for control in the process industries for small particle sizes using simple equipment. In this paper, size distributions obtained by ultrasonic attenuation spectroscopy (UAS). laser diffraction spectroscopy (LDS), focused-beam reflectance measurement (FBRM), and microscopic image analysis (MIA) are compared and the influence of crystal shape on size measurements is investigated. Data obtained from UAS are combined with information on shape factor (circularity) from imaging to obtain equivalent crystal-size distribution (CSD). Data obtained from LDS are converted with a shape factor (sphericity) from imaging to obtain CSD. In this paper, the influence of shape on size measurement is observed for these organic and inorganic chemicals.

Items from the Neighborhood Cohesion Instrument were completed by 1,732 individuals from a random sample of households in 20 rural communities across Canada during the summer of 2001. Confirmatory factor analysis of the NCI items based on polychoric correlations and weighted least squares estimation found three underlying latent variables. Although items were related to the three latent variables in a somewhat different manner than they were in Buckner’s original study, the same three latent variables were evident, providing convincing evidence that social cohesion has at least three subscales: psychological sense of community, neighboring, and attraction. Correlations between subscales were relatively high (between .67 and .87). Intraclass correlation coefficients for the three scales were .115, .127, and .112. In addition, the community means differed on different subscales in a manner that related to recognized characteristics of the communities. Thus the subscales are appropriate measures for both individuals and communities and can be recommended for further research on social cohesion.

The ultrasonic attenuation (UA) technique was developed to measure not only particle size distribution (PSD) but also concentration for on-line and off-line analysis of high concentration dispersions. Another significant characteristic along with particle size is particle shape. Particle shape influences the apparent particle size as measured by various instruments. Shape factors have been calculated from microscopic images using conventional image analysis (IA) techniques. Particle sizing data measured by UA is more representative than by IA because much larger samples are used for analysis. In this study UA, laser diffraction (LD) and IA size measurements are compared and the influence of particle shape is investigated.Data obtained from UA has been combined with information on shape factor from imaging to obtain equivalent particle size distributions. In a previous study [1], the influence of shape on size measurement has been investigated using different sizing techniques for non-fragile materials. These particles showed significant variation in measured PSD with their different shapes. In this paper, the same phenomenon was observed with evolving polymorphs of L-glutamic acid crystals (alpha and beta form), oxalic acid, sucrose and monosodium glutamate.

The Lasentec focused-beam reflectance measurement (FBRM) is becoming a more popular technique to measure particle size on-line in different applications. The FBRM uses a focused beam of laser light that scans across particles passing in front of the probe window to measure a chord length distribution (CLD). Compared with CLD information, the particle size distribution (PSD) is more useful because it is directly related to product quality and process productivity. However, it is not straightforward to convert a measured CLD into its corresponding PSD accurately due to the lack of a theoretical analysis for non-spherical particle systems. In this paper, firstly a general model to translate a PSD into its corresponding CLD is given for different shapes including spherical, ellipsoidal and more general non-spherical particles. Then an iterative inversion method is developed to obtain the PSD from a measured CLD. Finally, effectiveness of the proposed PSD-CLD model and iterative inversion method has been extensively validated by experiments.

In this paper, the theory on the translation of a measured chord length distribution (CLD) into its particle size distribution (PSD), which was developed in the first part of this study [Li and Wilkinson, 2005. Determination of non-spherical particle size distribution from chord length measurements. Part 1: theoretical analysis. Chemical Engineering Science 60, 3251–3265], has been validated using experimental results. CLDs were measured using the Lasentec focused beam reflectance measurement (FBRM) with three different materials, spherical ceramic beads and non-spherical plasma aluminium and zinc dust particles. Meanwhile, the particle shape and PSD of each material were also investigated by image analysis (IA). Comparison of the retrieved PSDs with the measured PSDs by IA shows that the PSD can be retrieved from a measured CLD successfully using the proposed iterative nonnegative least squares (NNLS) method based on the PSD–CLD model.

In this paper, scale up effects have been investigated for three geometrically similar laboratory scale vessels of 0.5, 2 and 20l with retreat curve impellers and cylindrical baffles, which mimic reactors widely used in the pharmaceutical and fine chemical industries, using CFD simulations. The convergence of fully three-dimensional, time-dependent numerical computations has been monitored to ensure the simulations reached quasi-steady state. CFD results have then been validated using LDA measurements and empirical power consumption literature data. The comparisons of power number, discharge flow number, secondary circulation flow number and pumping efficiency at three different scales suggest that the selection of scale of a laboratory vessel has little effect on the macro mixing performance for optimisation of the configuration and operating conditions of an industrial scale reactor.

In this article, particle size distributions (PSDs) measured by different techniques, including image analysis (IA), laser diffraction (LD), ultrasonic attenuation spectroscopy (UAS), and focused-beam reflectance measurement (FBRM), are compared for spherical glass beads and nonspherical silica flakes. It is shown that particle shape strongly affects the results obtained by different techniques. For spheres, the PSDs obtained by IA, LD, and UAS agree well. There is no consistent result among different particle measurement techniques for nonspherical particles. The conversion between PSDs obtained by IA, LD, and UAS has been based on particle shape factors. Caution must be exercised when a measured chord length distribution (CLD) is used to indicate the PSD during a process because the CLD result obtained by FBRM is complex, depending not only on the PSD, but also on particle optical properties and shape.

In this paper, extensive theoretical studies are described on two important issues in translating a chord length distribution (CLD) measured by FBRM instrument into its particle size distribution (PSD) including PSD–CLD and CLD–PSD translation models for general non-spherical particles. Analytical solutions to calculate the PSD–CLD models for spherical and ellipsoidal particles are developed. For non-spherical particles, a numerical method is given to calculate the PSD–CLD model. The iterative non-negative least squares (NNLS) method is proposed in the CLD–PSD model, because of its many advantages converting measured CLD into its PSD, such as insensitivity to measurement noise and particle shape. The effectiveness of the proposed methods is validated by extensive simulations.

Variable-temperature high-resolution capillary-mode powder X-ray diffraction is used to assess changes in unit-cell dimensions as a function of temperature over the range 188–328 K. No evidence was found for any polymorphic transformations over this temperature range and thermal expansion coefficients for urea were found to be αa= (5.27 ± 0.26) × 10−5 K−1and αc= (1.14 ± 0.057) × 10−5 K−1.

The effects of reactor internals and reactant mixing on the measured metastable zone width (MSZW) associated with the batch crystallization of l-glutamic acid from supersaturated aqueous solutions are presented. The results of cooling crystallization experiments, as carried out at three reactor scales (450 mL, 2 L, and 20 L) agitated at various stirring speeds using an industry-standard retreat curve impeller with a single beaver-tail baffle, are shown. The observed MSZWs are mostly found to decrease with increasing stirring speed, with enhanced nucleation also being observed as the reactor scale increased; albeit hindered nucleation was found at higher stirrer speeds in the 450-mL reactor experiments. The MSZW data are correlated with Reynolds number to reveal a model reflecting the combined influences of hydrodynamics and scale on the overall nucleation process.

LDA measurements and CFD predictions of the flow in a vessel stirred by a retreat curve impeller are reported. The CFD simulation was carried out using a commercial code CFX 5.5.1. The computational results have been extensively validated through the phase-resolved and phase-averaged LDA measurements to avoid the errors associated with pseudo-turbulence. It is shown that the axial and radial velocities were well predicted quantitatively over the whole vessel by the CFD simulation but the predicted tangential and turbulent kinetic energy are less close in agreement with the experimental data.

In situ ultrasonic attenuation spectroscopy is applied to the challenging case of monitoring the nucleation and growth of copper sulfate pentahydrate crystallized from supersaturated aqueous solutions, a system not readily amenable to analysis via optical methods due to the intense blue color of the saturated crystallizing solution. In experiments in the 2.8-L rectilinear reactor of a spectrometer, crystallization and dissolution points are reliably detected from the measured attenuation spectra. There are minor differences between data taken at 10 and 50 MHz, notably the lower frequency data appearing to be more sensitive to the particle formation/dissolution process. The nucleation data reveal that the material crystallizes fairly easily as characterized by a metastable zone width of ca. 3-4 degreesC and a significantly cooling-rate-dependent nucleation order of reaction of ca. 1.7 reflecting the fact that for high cooling rates the nucleation rate is less than that associated with supersaturation generation. The evolving crystal size distribution following nucleation, calculated from ultrasonic attenuation spectroscopy measurements, reveals well-defined oscillations in the observed crystal sizes consistent with the break-up of crystals larger than ca. 250 mum in this crystallizer. From dynamic measurements of changing particle size and concentration during the crystallization process, apparent crystal mass growth rates are calculated to be between 2.3 x 10(-3) kg/m(2).s at the maximum cooling rate of 0.55 degreesC/min and 2.02 x 10(-4) kg/m(2).s at the minimum cooling rate of 0.2 degreesC/min. On the basis of these data and assuming a spherical particle model, the linear crystal growth rates are estimated to be between 2.0 x 10(-6) m/s at the maximum cooling rate of 0.55 degreesC/min and 1.8 x 10(-7) m/s at the minimum cooling rate of 0.2 degreesC/min.