Measurement Of Bubble Size Distribution Research Articles

Broadband attenuation-based measurements of bubble size distributions

Oceanic bubbles generated by breaking waves are a key mechanism for air-sea gas exchange. Dense bubble plumes can also interfere with acoustic propagation in shallow water due to their high scattering and attenuation. Most studies of oceanic bubbles have used narrow-band acoustical and optical techniques to quantify bubble size distributions under breaking waves in open water. To study bubble distributions at estuarine tidal fronts, where localized downwelling and advection can generate large bubbles plumes, we developed a measurement system for estimating bubble size distributions via broadband excess attenuation. The system consists of two transducers (3–30 kHz and 30–110 kHz) and two hydrophones mounted on a towable frame. Testing was performed in a laboratory wave tank to measure bubble distributions below a field-scale (1 + meter) breaking wave. This presentation will discuss the bubble size distributions observed at different locations along the breaking wave crest in the direction of propagation and their associated time scales and sound speed anomalies. We also compare these distributions with those previously observed in open water and will present preliminary results from a field deployment of the system in the Connecticut River estuary.

Bubble size analysis in a two-phase countercurrent flow in the narrow rectangular column

The flow of bubbles in a two-phase system has great implications in chemical, petrochemical, and biochemical applications. This work enunciates the measurement of bubble size distribution and bubble aspect ratio in three-different axial zones in the countercurrent flow mode with a gas and liquid velocity range of 0.044–0.321 and 0.019–0.058 m/s, respectively. Bubble size is measured by the photographic technique. The bubble aspect ratio and bubble size distribution results reveal that the impact of gas and liquid velocities is significant on the Sauter mean bubble size. The Sauter mean bubble size increases as the gas velocity increases, whereas it decreases with the liquid velocity. The Sauter mean bubble diameter ranges from 2.65 to 6.16 mm. The distribution of bubble sizes follows the LogLogistic probability density function. In addition, a correlation is also proposed for the interpretation of bubble diameter in terms of Reynolds number and Froude number. The bubble aspect ratio changes with axial zones and gas and liquid velocities. Experiments reveal that the bubble aspect increases with liquid velocity while decreasing with gas velocity and axial zones. The bubble aspect ratio correlations are developed in terms of Eötvös and Reynolds numbers. The present results will be helpful for the process intensification of bubble-aided two-phase flow applications.

Influence of counterdiffusion effects on mass transfer coefficients in stirred tank reactors

Accurate knowledge of volumetric oxygen mass transfer coefficients in multiphase systems is important for successfully designing and efficient operation of process plants. As a result, a large number of papers on the determination of volumetric mass transfer coefficients kLa have been published in the literature. The dynamic degassing method is probably the most common method for determining the volumetric mass transfer coefficient. However, little work is known on the influence of countercurrent diffusion effects, arising from further dissolved components, on the volumetric mass transfer coefficient.For this reason, the effect of countercurrent diffusion is investigated in the present work by using different stripping gases to determine the volumetric oxygen mass transfer coefficient. Furthermore, the effect of counterdiffusion is investigated for both conventional aeration and aeration with microbubbles.The present work shows that the choice of stripping gas has a decisive influence on the determination of the volumetric oxygen mass transfer coefficient. Moreover, it can be shown that this effect is directly proportional to the solubility of the stripping gas in the aqueous phase. Simultaneous measurements of bubble size distributions allow determining mass transfer coefficients kL. A model is developed describing the decrease in the mass transfer coefficient as a function of the solubility of the stripping gas.Furthermore, it can be shown that for the experimental determination of volumetric oxygen mass transfer coefficients, the choice of stripping gas should be adapted to secondary gas types occurring within real processes achieving a better comparability.

Vortex-generator-induced intensification of fine mineral collection in pipe flow

Flotation of fine minerals is becoming increasingly challenging. Mineral particles are believed to collide and attach efficiently to air bubbles in highly turbulent circular pipes. In this study, vortex generators (VGs) are adopted to strengthen the turbulence and improve the collection efficiency of fine minerals in pipe flow flotation. Computational fluid dynamics simulations, bubble size measurements, and flotation tests are conducted in a modified pipe unit with VGs and an initial pipe unit without VGs. The simulation results show that the VGs induce streamwise vortices and increase the turbulent dissipation rate in the pipe unit. Bubble size distribution measurements show that the VGs decrease the bubble size slightly at high velocities. Flotation tests prove that vigorous turbulence induced by VGs is beneficial to the recovery of fine particles. In particular, the flotation rate constant of −20 µm particles at a liquid velocity of 0.7 m/s increases by more than 40% in the presence of VG-s5. Nevertheless, the excessive turbulence induced by VG-s5 caused a 20% decrease in the flotation rate of +45–74 μm particles when the liquid velocity increases to 2.0 m/s. This indicates that a relationship exists between turbulence and the particle size recovered. For a specified particle size, turbulence can be adjusted by varying the vortex generator spacing and changing the flow velocity to achieve the optimum flotation performance.

Bubble size distributions in gas–liquid–solid systems and their influence on flotation separation in a bubble column

Measurements of bubble size distributions are very important in flotation which is widely used in different chemical industries such as the recovery of valuable minerals, wastewater treatment, deinking of waste paper, and oil recovery from tar sands. Although numerous authors determined bubble size distributions (BSD-s) in gas–liquid systems, BSD-s have been rarely determined in gas–liquid–solid systems. Moreover, the influence of BSD-s on flotation has not been quantified under a wide range of experimental conditions. Thus, this paper is focused on measurements of bubble size distributions by changing surfactant dosage, solid percentage, air velocity and particle size as well as establishing the relationships between bubble size distribution properties and flotation performances. It was found that the experimental bubble size distributions are successfully fitted by using the log-normal distribution model in which the two model parameters (μ and σ) are used. The results showed that the most important factors affecting the Sauter mean diameter (d32), the model parameters and the flotation recoveries are the surfactant dosage and the solid percentage. The increase in the surfactant dosage decreased d32 probably due to the decrease in liquid surface tension. The increase in solid percentage resulted in an increase in d32 perhaps due to the increase in the pulp viscosity at high solid percentage. The lower the d32, the lower the flotation recovery. The results show that d32 and the model parameters represent the true BSD-s. However, a thorough understanding of bubble size distributions is needed to better understand flotation performances.

Inversion method of bubble size distribution based on acoustic nonlinear coefficient measurement**Project supported by the National Natural Science Foundation of China (Grant Nos. 11674074 and 61701133).

Measurements of bubble size distribution require the understanding of the acoustic characteristics of the medium. The bubbles show highly nonlinear properties under finite amplitude acoustic excitation, so the acoustic fields from bubble population are easily observed at the second harmonics as well as at the fundamental frequency, which shows that the nonlinear coefficient increases obviously. The inversion method of bubble size distribution based on nonlinear acoustic effects can peel off the influence of complex environment and obtain the size distribution coefficient information of bubbles more accurately. The previous nonlinear inversion methods of bubble size distribution are mostly based on the nonlinear scattering cross-section characteristics of bubbles. However, the stability of inversion is not high enough. In this paper, we introduce a new acoustic inversion method for bubble size distribution, which is based on the nonlinear coefficients of bubble medium. Compared with other inversion methods based on linear or nonlinear scattering cross section, the inversion method based on nonlinear coefficients of bubble medium proposed in this paper shows good robustness in both simulation and experiment.

Determination of Bubble Size Distribution Using Ultrasound Array Imaging.

In this article, ultrasonic phased arrays are deployed as an imaging tool for industrial process analysis. Such arrays are typically used for sonar, medical diagnosis, and nondestructive testing; however, they have not yet been applied to industrial process analysis. The precise positioning of array elements and high frequencies possible with this technology mean that highly focused images can be generated, which cannot currently be achieved using ultrasound tomography. This article aims to highlight the potential of this technology for the measurement of bubble size distribution (BSD) and to demonstrate its application to both intrusive and noninvasive process measurements. Ultrasound images of bubble reflectors are generated using the total focusing method deployed using a 32-element, 5-MHz linear phased array, and an image processing algorithm for BSD determination is presented and evaluated under stationary and dynamic acquisition conditions. It is found that the sizing accuracy is within 10% for stationary reflectors larger than 4λ in diameter and that the algorithm is stable across the expected spatial variation of reflectors. The phased array is coupled to a six-axis robotic arm to scan a solid sample containing bubble reflectors at velocities up to 500 mms-1. The sizing accuracy is within 45% for bubbles larger than 4λ in diameter and at velocities up to 300 mms-1. However, above this velocity, the algorithm breaks down for reflectors smaller than 9λ in diameter. The ultrasound system is applied to a stream of air bubbles rising through water, which is verified via photographic analysis. Images were generated both intrusive and noninvasive, via a 10-mm Perspex barrier, to the process stream. The high bubble density in the process stream introduced scattering, limiting the measurement repeatability and the sample size in the measured distribution. Notwithstanding, this result demonstrates the potential of this technology to size bubbles for intrusive and noninvasive process analyses.

Experimental study of external characteristics of a flashing water jet in water

External characteristics of a flashing water jet released in water were studied using Particle Image Velocimetry and Shadowgraphy techniques for different pressure ratios and nozzle diameters. The flow dynamics of such a jet are dominated by vapor bubble formation due to thermal non-equilibrium and phase change processes. In order to gain insights into the physics of such complex flows, experiments were performed and measurements on velocity distribution and turbulence statistics were obtained. The bubble size distributions were also recorded to characterize the bubble disintegration mechanism. Flashing jet emanating from two different nozzle diameters, D = 5 mm and 2 mm, and at three different pressure ratios, PR = 1.5 bar, 2 bar and 2.5 bar, were studied, at a fluid inlet temperature of To = 380 K. The results revealed that the maximum value of centerline velocity occurs far downstream, at x ≥ 0.02 m from the nozzle exit, due to the presence of vapor bubble formation that leads to an interaction between the liquid and gaseous phases. This is in contrast to a canonical water jet (i.e., without phase change), where the centerline velocity attains a maximum at the nozzle exit. For D = 5 mm, the turbulent kinetic energy (TKE) was higher and the decay rate was slower as compared to D = 2 mm for PR ≥ 2 bar. From the centerline velocity and TKE measurements, a critical injection pressure ratio, PR > 1.5 bar was documented, beyond which the external flow dynamics of the jet is governed primarily by the flashing mechanism. These observations were verified from the bubble size distribution measurements and can be attributed to thermal non-equilibrium and phase change process. Lastly, the presence of coherent structures in the flow reveals that the entrainment dynamics of a flashing jet is dominated by vorticity.

Experimental measurements of bubble size distributions in a water model and its influence on the aluminum kinetics degassing

ABSTRACTAn experimental study of the gas‐liquid dynamics in a water model of an aluminum ladle was conducted. The rotor degassing performance was evaluated for two commercial rotor‐injector devices compared to a new rotor design. In this work, the influence of the turbulent properties of the flow fields on the bubble size distribution is analyzed for a better understanding of its impact on the degassing efficiency in aluminum refinement operations. The degassing process was analyzed by two different methods: (a) a high‐speed camera was used to obtain the bubble size distribution into the container; and (b) the particle image velocimetry technique (PIV) was employed to obtain the liquid flow properties. It was found that the rotor geometry plays an important role on the average size and distribution of the bubbles. The energy dissipation rate contours show significant differences for the distinct rotors tested. These hydrodynamic states and bubbles distribution dominate the kinetic and efficiency of the degassing processes. It was shown that the new rotor configuration enhances the degassing kinetics, due to a more suitable bubble dispersion compared to the commercial rotors tested.

Measurement of bubble size distribution in activated sludge bubble column bioreactor

Bubble size, bubble rise velocity and the corresponding distributions were directly measured by a dual-tip electro resistivity probe and to present empirical correlations to estimate the average bubble size and bubble rise velocity in activated sludge. Experiments were conducted in a 9cm diameter bubble column while superficial gas velocity was in range of 0.5–7cm/s. Four samples of activated sludge with mixed liquor suspended solid (MLSS) concentrations from 1.082 to 10.82g/L were used in the experiments. It was found that increasing the MLSS concentration resulted in increase in the bubble chord length and bubble rise velocity and decrease in the gas hold up. It was shown that in the homogeneous regime, the probability distributions of bubble chord length and bubble rise velocity are sharp and unimodal. To estimate bubble size and bubble rise velocity in activated sludge, correlations were developed which includes effect of MLSS concentration besides physical properties and operating conditions. The experimental results of this study were compared with the proposed correlations in literature obtained in two and three phase bubble column. It was shown that neither of the previous correlations can predict the bubble size and bubble rise velocity in the activated sludge system properly while the proposed correlation in this work shows a good precision for estimating these parameters.

An integrative image measurement technique for dense bubbly flows with a wide size distribution

The measurements of bubble size distribution are ubiquitous in many industrial applications in chemical engineering. The conventional methods using image analysis to measure bubble size are limited in their robustness and applicability in highly turbulent bubbly flows. These flows usually impose significant challenges for image processing such as a wide range of bubble size distribution, spatial and temporal inhomogeneity of image background including in-focus and out-of-focus bubbles, as well as the excessive presence of bubble clusters. This article introduces a multi-level image analysis approach to detect a wide size range of bubbles and resolve bubble clusters from images obtained in a turbulent bubbly wake of a ventilated hydrofoil. The proposed approach was implemented to derive bubble size and air ventilation rate from the synthetic images and the experiments, respectively. The results show a great promise in its applicability for online monitoring of bubbly flows in a number of industrial applications.

Gas Bubble Size Measurements in Liquid Mercury Using an Acoustic Spectrometer

A properly dispersed population of small bubbles can mitigate cavitation damage to a spallation neutron source target. In order to measure such a bubble population, an acoustic device was developed and implemented in a mercury loop at ORNL. The instrument generated pulses of various frequencies and measured their acoustic propagation in the bubbly medium. It then deduced sound speed and attenuation at the various frequencies and used an inverse problem solver to provide near real-time measurements of bubble size distribution and void fraction. The measurements were then favorably compared with an optical method.

Investigation of Bubble and Fluid Flow Patterns Within a Column Photobioreactor

This research study investigates bubble and liquid circulation patterns in a vertical column photobioreactor (PBR) both experimentally as well as computationally using computational fluid dynamics (CFD). Dispersed gas-liquid flow in the rectangular bubble column PBR are modeled using Eulerian-Lagrangian approach. A low Reynolds number k-epsilon CFD model is used to describe the flow pattern near the wall. A flat surface bubble column PBR is used to achieve sufficient light penetration into the system. Bubble size distribution measurements were completed using a high-speed digital camera. Operating parameters, bubble flow patterns, and internal hydrodynamics of a bubble column reactor were studied, and the numerical simulations presented for the hydrodynamics in a bubble column PBR account for bubble phenomena that have not been sufficiently accounted for in previous research. Bubble size and shape affect the hydrodynamics as does bubble interaction with other bubbles (multiple bubbles in a flow versus single bubbles and wall effects on bubble(s) that are not symmetrical or bubbles not centered on the reactor cross-section). Understanding the bubble movement patterns will aid in predicting other design parameters like mass transfer (bubble to liquid and liquid to bubble), heat transfer (within the PBR and between the PBR and environment surrounding the PBR), and interaction forces inside the PBR. The computational results are validated with experimental data and from current literature.

Applications of ultra-fast MRI to high voidage bubbly flow: Measurement of bubble size distributions, interfacial area and hydrodynamics

Ultrafast magnetic resonance imaging (MRI) has been applied to high voidage bubbly flow systems for the first time. Using an MRI protocol selected for fast acquisition (15.3ms per image) and robustness to fluid shear, images were obtained of bubbly flow in a vertical pipe of internal diameter 31mm for systems up to a gas voidage of 41%. From these images, two measurements of bubble size were extracted (one on the basis of volume and one on the basis of projected radius), which allowed the quantification of both the bubble size distribution (BSD) and bubble shape (and hence interfacial area). These measurements were found to be accurate to within 5.6% for voidages up to 22%. Measurements were acquired at 10 positions along a column for an electrolyte stabilised system to show the evolution of the BSD. Additionally, using a velocimetric variant of the MRI technique, maps of the fluid velocity in the vertical direction were obtained for all voidages. High temporal resolution series of MRI images were also acquired, from which it is possible to track the motion of individual bubbles. These measurements are non-invasive and applicable to optically opaque systems. The MRI technique should prove useful both for phenomenological studies of high voidage bubbly flow, and for the validation of models of multiphase flow systems.

New image-processing algorithm for measurement of bubble size distribution from flotation froth images

Research and experience have demonstrated that the operating conditions of the flotation process are reflected in the froth’s appearance. Despite recent advances in image analysis and several algorithms developed for froth flotation, there still is not a comprehensive algorithm to accurately determine bubble features from actual froth images. Being able to accurately and automatically measure bubble size distribution is an important requirement for optimization and control of the flotation process. Segmentation is one of the most effective approaches for accurately determining the bubble size distribution. In the present study, a new bubble segmentation algorithm, which utilizes an adaptive marker-based watershed transform, is proposed to measure the bubble size distribution from froth images. The developed algorithm is validated using some industrial scale froth images from flotation cells at different duties.

On the Carrying Capacity Limitation in Large Flotation Cells

Mineral carrying rates across the pulp-froth interface in the range of 1.4-3.2 tph/m2 and bubble surface coverage between 7 and 22 %, have been determined in large rougher flotation cells (100, 160 and 300 m3). A model for estimating the carrying rate at the pulp-froth interface as a function of bubble surface coverage, bubble surface area flux and Sauter mean particle size of the mineral collected by true flotation was developed. To evaluate these variables, measurements of bubble size distribution, bubble load and size distribution of collected particles were performed at different concentrators. The results of bubble surface coverage were significantly lower than the expected maximum coverage for the collection zone, indicating that for copper rougher flotation cells it is unlikely to achieve maximum carrying rate at the pulp-froth interface level. The main constraints regarding carrying capacity limitations at the concentrate overflow were then related to the froth transport characteristics such as froth recovery.RésuméOn a déterminé les vitesses de transport de minéral à travers l'interface pulpe-mousse dans la gamme de 1.4-3.2 tph/m2 et entre 7 et 22% de couverture de la surface de bulle, dans des cellules de flottation de dégrossissage de grande taille (100, 160 et 300 m3). On a développé un modèle pour estimer la vitesse de transport à l'interface pulpe-mousse en fonction de la couverture de la surface de bulle, du flux de la surface de bulle et de la taille moyenne de Sauter des particules de minéral captées par flottation véritable. Afin d'évaluer ces variables, on a effectué, à différents concentrateurs, des measures de la distribution de la taille de bulle, de la charge de bulle et de la distribution de taille des particules captées. Les résultats de la couverture de la surface de bulle étaient significativement plus faibles que la couverture maximale attendue pour la zone de captage, indiquant que pour les cellules de flottation de dégrossissage du cuivre, il était peu probable d'obtenir la vitesse maximale de transport au niveau de l'interface pulpe-mousse. On a ensuite relié les contraintes principales ayant trait aux limitations de la capacité de transport au déversement du concentré, aux caractéristiques de transport de la mousse comme la récupération de la mousse.

Development of an acoustic instrument for bubble size distribution measurement

Measurement of bubble size distribution and void fraction is of vital importance in many multi-phase flow applications. This paper describes an acoustics based device, the ABS Acoustic Bubble Spectrometer®, which can conduct measurements accurately in near real-time in a cost-effective fashion. By propagating short bursts of sound at different frequencies through bubbly medium, it measures frequency dependent attenuations and phase velocities of the acoustic waves and uses them to obtain the bubble size distribution (number of bubbles per size) by solving an inverse problem. Recent developments, both in hardware and software, as well as their validations are presented, these new advancements enable the ABS to measure void fractions up to 3×10−3 with bubble sizes ranging from 10 μm to 3mm.

Measurement of bubble size distribution in a gas–liquid foam using pulsed-field gradient nuclear magnetic resonance

Pulsed-field gradient nuclear magnetic resonance, previously used for measuring droplet size distributions in emulsions, has been used to measure bubble size distributions in a non-overflowing pneumatic gas-liquid foam that has been created by sparging propane into an aqueous solution of 1.5g/l (5.20mM) SDS. The bubble size distributions measured were reproducible and approximated a Weibull distribution. However, the bubble size distributions did not materially change with position at which they were measured within the froth. An analysis of foam coarsening due to Ostwald ripening in a non-overflowing foam indicates that, for the experimental conditions employed, one would not expect this to be a significant effect. It is therefore apparent that the eventual collapse of the foam is due to bubble bursting (or surface coalescence) rather than Ostwald ripening. This surface coalescence occurs because of evaporation from the free surface of the foam. An analytical solution for the liquid fraction profile for a certain class of non-overflowing pneumatic foam is given, and a mean bubble size that is appropriate for drainage calculations is suggested.

Experimental analysis of bubble size distributions in 2D gas fluidized beds

The present work focuses on the development of a novel statistical approach for the analysis of the behavior of a 2D fluidized bed operating under bubbling and slugging conditions. Experimental data were collected by means of a purposely built lab-scale 2D fluidized bed and an original digital image analysis technique (DIAT), allowing the acquisition of a number of data sufficient to perform a statistical analysis of bubble dynamics. In particular, measurements of bubble size distributions (BSD) along the bed were performed. The analysis allowed to characterize the shape of BSD for different particle systems and fluidization gas velocities; interestingly a bimodal shape of BSD is observed, likely due to the simultaneous presence of coalescence and splitting phenomena. Linear combination of two Gamma distributions was used to describe the data. Also bubble size distributions weighed on bubble area were computed, and for all cases a characteristic negatively skewed distribution shape was found. Preliminary assessment of available data allowed to point out the effectiveness of the new approach proposed for a more in-depth description of bubble dynamics in 2D gas fluidized beds.

On the measurement of bubble size distribution in gas–liquid contactors via light sheet and image analysis

Abstract Particle image velocimetry techniques coupled with advanced image processing tools are receiving an increasing interest for measuring flow quantities and local bubble-size distributions in gas–liquid mechanically agitated vessels. When trying to analyze image information the problem arises that bubble sizes are generally underestimated, due to the fact that the laser sheet used for lighting the system randomly cuts bubbles over non-diametrical planes, leading to an apparent bubble size distribution even in the ideal case of single sized bubbles. Clearly in the case of bubbles with a size distribution the experimental information obtained is affected by the superposition of effects. Aim of this work is that of providing a numerical procedure able to reconstruct actual bubble size distributions from relevant apparent size distributions obtained by laser sheet illumination and image analysis. The procedure proposed is robust and viable and can account for laser sheet thickness. The procedure is shown to provide fully satisfactory results even with quite extreme distributions. BSD resolution dependence on the numerousness of raw data processed is discussed. Use of the proposed procedure for extracting BSD from bubble chord raw data obtained by other devices, such as point probes, is straightforward.