Bubble Edge Research Articles

Arc bubble edge detection method based on deep transfer learning in underwater wet welding.

The stability of arc bubble is a crucial indicator of underwater wet welding process. However, limited research exists on detecting arc bubble edges in such environments, and traditional algorithms often produce blurry and discontinuous results. To address these challenges, we propose a novel arc bubble edge detection method based on deep transfer learning for processing underwater wet welding images. The proposed method integrates two training stages: pre-training and fine-tuning. In the pre-training stage, a large source domain dataset is used to train VGG16 as a feature extractor. In the fine-tuning stage, we introduce the Attention-Scale-Semantics (ASS) model, which consists of a Convolutional Block Attention Module (CBAM), a Scale Fusion Module (SCM) and a Semantic Fusion Module (SEM). The ASS model is further trained on a small target domain dataset specific to underwater wet welding to fine-tune the model parameters. The CBAM can adaptively weight the feature maps, focusing on more crucial features to better capture edge information. The SCM training method maximizes feature utilization and simplifies training by combining multi-scale features. Additionally, the skip structure of SEM effectively mitigates semantic loss in the high-level network, enhancing the accuracy of edge detection. On the BSDS500 dataset and a self-constructed underwater wet welding dataset, the ASS model was evaluated against conventional edge detection models-Richer Convolutional Features (RCF), Fully Convolutional Network (FCN), and UNet-as well as state-of-the-art models LDC and TEED. In terms of Mean Absolute Error (MAE), accuracy, and other evaluation metrics, the ASS model consistently outperforms these models, demonstrating edge detection capabilities that are both effective and stable in detecting arc bubble edges in underwater wet welding images.

Electron Dynamics in a Relativistic Central Field Approach Associated with Laser Acceleration of Electrons

We perform 2D PIC simulations and observe different scenarios of electron trajectories trapped inside the bubble formed during the laser wakefield acceleration process (LWFA). The trajectories can close on the inside or on the outside of the bubble. The evolution of maximum energy of the injected electrons in both cases is evaluated. We propose an analytic approach based on a relativistic central field model for describing the attraction of electrons near the bubble edge and compare it with the simulations.

Fermi-bubble Bulk and Edge Analysis Reveals Dust, Cooling Breaks, and Cosmic-Ray Diffusion, Facilitating a Self-consistent Model

The full, radio to γ-ray spectrum of the Fermi bubbles is shown to be consistent with standard strong-shock electron acceleration at the bubble edge, without the unnatural energy cutoffs and unrealistic electron cooling of previous studies, if the ambient interstellar radiation is strong; the γ-ray cooling break should then have a microwave counterpart, undetected until now. Indeed, a broadband bubble-edge analysis uncovers a pronounced downstream dust component, which masked the anticipated ∼35 GHz spectral break, and the same overall radio softening consistent with Kraichnan diffusion previously reported in γ-rays. A self-consistent bulk and edge model implies a few Myr old bubbles, with fairly uniform radiation fields and enhanced magnetization near the edge.

3D Simulations of Nucleate Boiling with Sharp Interface Vof and Localized Adaptive Mesh Refinement in Ansys-Fluent

Abstract The present work demonstrates the use of customized Ansys-Fluent in performing 3D numerical simulations of nucleate boiling with a sharp interface and adaptive mesh refinement. The developed simulation approach is a reliable and effective tool to investigate 3D boiling phenomena by accurately capturing the thermal and fluid dynamic interfacial vapor-liquid interaction and reducing computational time. These methods account for 3D sharp interface and thermal conditions of saturation temperature refining the mesh around the bubble edge. User-Defined-Functions (UDFs) were developed to customize the software Ansys-Fluent to preserve the interface sharpness, maintain saturation temperature conditions, and perform effective adaptive mesh refinement in a localized region around the interface. Adaptive mesh refinement is accomplished by a UDF that identifies the cells near the contact line and the liquid-vapor interface and applies the adaptive mesh refinement algorithms only at the identified cells. Validating the approach considered spherical bubble growth with an observed acceptable difference between theoretical and simulation bubble growth rates of 10%. Bubble growth simulations with water reveal an influence region of 2.7 times the departure bubble diameter, and average heat transfer coefficient of 15000 W/m2-K. In addition, the results indicate a reduced computational time of 75 hours using adaptive mesh compared to uniform mesh.

SBSNet: selective branch segmentation network of coal flotation froth images

ABSTRACT Flotation is a physical and chemical process that uses the difference between mineral surface properties to separate valuable minerals from worthless minerals through the attachment of bubbles. The visual information of the flotation foam surface can reflect the flotation effect and will be closely related to the flotation conditions, directly indicating the degree of mineralization of the foam layer. Considering the problem that coal slurry foam images are difficult to segment due to the mutual adhesion of bubbles and fuzzy boundaries, in this work, we proposed a new and effective segmentation algorithm for extracting coal slurry flotation foam edge information. Based on the encoder and decoder structure, a selective multi-branch input segmentation deep neural network model was designed. We evaluated the proposed method and model using a self-built coal slurry foam image segmentation dataset. The intersection over union, mean intersection over union, and PA evaluation indicators of this method increased by 12.8376%, 7.2716%, and 1.457%, respectively, compared with those of the U-Net model. The experimental results showed that the selective branch segmentation network model had high accuracy and robustness in extracting coal slurry bubble edge information, and it could effectively solve the problem of difficult segmentation of coal slurry foam images. This study provided a foundation for optimizing flotation process parameters and realizing intelligent flotation site parameters.

The impacts of material acoustic impedance and thickness on single laser-induced bubble dynamics and determining factors in resulting pressure

The objective of this paper is to experimentally identify the primary sources of pressure when a laser-induced cavitation bubble is collapsing to a wall with specific emphases on the material acoustic impedance and thickness. Both high-speed videos and local wall pressure measurements were performed for various standoff ratios γ, bubble diameters, and wall materials. In the case of a rigid wall, in addition to the known high pressure for γ<0.6 where the bubble attaches and collapses on the wall (ring collapse), at γ≈1.12 where the jet is dominant, and low pressure obtained at γ≈0.913, where neither effect is significant, we further captured similar pressure profiles during the collapse after the first rebound at γ≈1.16 for the ring collapse, γ≈1.79 for the jet, and γ≈1.41 for the minimal, respectively. This indicates a strong jet is typically followed by a strong ring collapse. Generally, the pressure from the second collapse increases faster with the bubble size than that of the first collapse. For walls featuring smaller acoustic impedance or thickness, which cannot be approximated as rigid bodies or accessed by pressure sensor, our unique bubble edge analyzing tool shows that the ring collapse and jet effects are moved to smaller values of γ. The maximum pressure exerted on the wall in these cases is smaller than that on the rigid wall. Finally, we summarized the asymptotic evolution curves of each edge which bound the bubble dynamics at different standoff ratios.

NH3/O2 premixed combustion in a single bubble of fluidized bed

As a carbon - free, economical and safe energy carrier, ammonia is favoured and has been used in many different applications for combustion. However, there are few reports of NH3 being burned in a fluidized bed. This investigation of the premixed combustion of NH3 and O2 in a single bubble inside a fluidized bed was done both experimentally and with CFD-DEM simulations incorporating a skeletal mechanism of 98 elementary reactions. The results show that premixed combustion of NH3 and O2 occurred in a growing bubble. Quenching became evident at a bubble's edge. Combustion became unstable at low equivalence ratios. In addition, the equivalence ratio significantly affected the ignition delay, flame's propagation velocity, as well as the concentrations of gaseous species in the bubble and emulsion phases. Increasing bed temperature reduced the ignition delay. Nitrogen, as the background fluidising gas, exerted a negative impact on combustion, because of N2 diffusing from the emulsion to the bubble phase. The combustion of NH3/air in the bubble phase of a fluidized bed was difficult to happen in the absence of a catalyst or co-combustion with an inflammable fuel.

Study on bubble pulsation process of underwater explosion between parallel plates with various distances

This paper presents a numerical and experimental study of the dynamic behavior of underwater explosion bubbles between two parallel plates at different detonation distances. To simulate the strong nonlinear interaction between cavitation bubbles and the near-wall boundary, the coupled Eulerian-Lagrangian (CEL) method is used. The evolution of bubble morphology in the fluid domain is observed simultaneously, revealing the flow inside the bubble and the pulsation process under various visual angles that are difficult to observe in the experiment. In the experimental system, micro-equivalent explosives are used to generate shock waves, and the shock wave load during bubble pulsation and wall pressure during collapse phase are measured. High-speed cameras capture the bubble morphology evolution process and dynamic behavior at different detonation distances, which are then compared with the numerical results to verify their accuracy.The study focuses on the bubble morphology evolution process with dimensionless distance φ ranging from 0.42 to 1.71, and the collapse modes of bubbles in different forms are discussed in detail. The pulsation process of bubbles in the free field is verified, and it is found that when the distance between the two plates is less than the maximum radius of the bubble, the bubble collapses in the center between the two plates as the dimensionless distance r increases. Moreover, the Locating column has a large interference on the bubble movement. The bubble surface is divided into five layers of ring ligament with obvious characteristics. This paper investigates the dynamic behaviour of cavitation bubbles between two parallel plates at different plate spacings. The bubble undergoes a series of shape changes, from a columnar cake to a multi-layer ring with obvious edge tomography, and finally collapses into a number of minor bubbles, producing a cavitation bubble band connecting the two cavitation layer ligaments. When the plate spacing is between 1 and 1.5 times the maximum bubble radius, the bubble edge contacts the target plate, causing it to change from a “Hamburg” shape to an hourglass shape, tear into two sub-bubbles, and then produce two bubble jets pointing at opposite target plates. The vortex at the edge of the bubble influences the bubble's shape, resulting in an obvious corner and the formation of an annular bubble. When the plate spacing is greater than 1.5 times the maximum bubble radius, the bubble is torn into two sub-bubbles, and the bubble edge is smooth and does not contact the target plate during expansion. The numerical model accurately simulates the bubble pulsation and jet formation process, which is validated by experimental results. However, the experimental system has limitations in terms of image interval and viewing angle, which can be addressed by using the numerical model to simulate the bubble pulsation process at a higher resolution and a three-dimensional perspective. This study provides valuable insights into the strong nonlinear interaction between bubbles and target plates at different detonation distances. The accuracy of the HPB measurement system is also verified, and the numerical method used in this study is effectively simulates the pulsation characteristics of cavitation bubbles, including the formation of jets and annular bubbles. This work lays a solid foundation for further studies on cavitation bubble dynamics and their interaction with target plates, which has important implications for the design and analysis of underwater structures subjected to explosive loads.

Effect of Swirl Gas Injection on Bubble Characteristics in a Bubble Column

Swirling gas injection is a well-known technique to improve mass transfer in bubble columns. It can be used to create small bubbles with a high surface area-to-volume ratio, which is beneficial for mass transfer. Swirl gas injection can also be used to create a more uniform bubble size distribution and improve the mixing of gas and liquid in the column. This study aims to determine the impact of swirl gas injection on bubble properties, including bubble shape, size, and velocity. A bubble detection approach has been developed for quick and precise determination of bubble size distributions in gas-liquid systems. Advanced digital image processing, including edge detection and bubble edge recognition, is used in this method. The experiment is conducted in a bubble column at a height of 57 cm and 61 cm. The column had a ring sparger and was made of Plexiglas. Tap water was used as the liquid, while air from an air compressor was utilized as the gas phase. The shape, size, population, and velocity of the bubble are measured using a high-speed digital camera. According to this study, the average bubble size reduced as the impeller speed increased, while the population of bubbles increased when the sparger rotation speed increased from 30 to 150 rpm.

A 3D measurement method of bubbles based on edge gradient segmentation of light field images

Based on the light field imaging technology, a three-dimensional(3D) measurement method of bubbles through the use of edge gradient segmentation (TMBEGS) is proposed in this paper. From the characteristic that the bubble edge image has great gray gradient, the coordinates of the bubble edge pixels are determined. The depths of the edge pixels are further derived by evaluating the sharpness of the edge pixels in the refocused light field images. On this basis, the depth coordinate of the bubble is calculated by averaging the depths of all edge pixels, and the 3D measurement of the bubble is then achieved by combining the centroid coordinate and equivalent diameter of the bubble. Finally, experiments are carried out to evaluate TMBEGS on a calibration setup and a bubbling bed experimental device. Results indicated that the relative errors of the bubble diameter and depth by TMBEGS are less than 1.18 % and 5.93 %, which are better than 2.99 % and 17.51 % by the existing method, respectively. On the bubbling bed experimental device, the equivalent diameter of bubbles is increased by 5.69 % compared with the existing method, and the standard deviation of the bubble depth is less than 1.78 mm, which is 21.3 % lower than the existing method. With the increase of the bubble equivalent diameter, the bubble surface deformation intensifies under the effects of the bubble surface tension, viscous force, and inertia force, which affects the imaging quality of the bubble, and reduces the measurement accuracy of the bubble edge and depth. These results illustrated that TMBEGS has high measurement accuracy of the bubbles and is capable of charactering the bubbles in the gas–liquid two-phase flow system.

Illumination sensitivity analysis for the speck-based froth detection method using potash flotation machines

Machine vision methods are widely used in the industry. The paper considers the possibilities of using the machine vision technology to monitor potash ore flotation processes. It provides an overview of ready-made solutions for bubble edge detection. The widely used methods, however, are not suitable for the detection of cleaner flotation froth due to the specifics of its froth bed coloring and structure. A bubble detection method is described that is based on the distance between the specks generated by a point light source. The common algorithm is as follows: image conversion to grayscale, binarization, post-processing, and bubble separation using the ABCalgorithm. Particular attention is paid to establishing the proper binarization threshold, as it affects the quality of the final image significantly. A certain binarization threshold selection method is substantiated. The purpose of this work was to establish the sensitivity of the speck-based method for detecting the froth bed parameters to the binarization threshold. The study considers 20 experimental image sets obtained using different flotation machines, both with and without an additional light source. Binarization threshold selection profiles were generated for each frame. The respective profile analysis allowed identifying the best binarization thresholds for various types of flotation machines. As a result, specific methods for removing frames with low illumination and algorithms for determining the threshold binarization values for different types of flotation machines are proposed.

Direct Numerical Simulation of Bubble Formation Through a Submerged “Flute” With Experimental Validation

Abstract Direct numerical simulation (DNS) is often used to uncover and highlight physical phenomena that are not properly resolved using other computational fluid dynamics methods due to shortcuts taken in the latter to cheapen computational cost. In this work, we use DNS along with interface tracking to take an in-depth look at bubble formation, departure, and ascent through water. To form the bubbles, air is injected through a novel orifice geometry not unlike that of a flute submerged underwater, which introduces phenomena that are not typically brought to light in conventional orifice studies. For example, our single-phase simulations show a significant leaning effect, wherein pressure accumulating at the trailing nozzle edges leads to asymmetric discharge through the nozzle hole and an upward bias in the flow in the rest of the pipe. In our two-phase simulations, this effect is masked by the surface tension of the bubble sitting on the nozzle, but it can still be seen following departure events. After bubble departure, we observe the bubbles converge toward an ellipsoidal shape, which has been validated by experiments. As the bubbles rise, we note that local variations in the vertical velocity cause the bubble edges to flap slightly, oscillating between relatively low and high velocities at the edges.

Effect of Rear Wing on Time-Averaged Ground Vehicle Wake With Variable Slant Angle

Abstract The slant angle plays a crucial role in the flow property of hatchback ground vehicles. An optimum slant angle is obligatory for better handling the ground vehicles when fitted with a rear wing. In this regard, the variation of time-averaged flow properties around a wing-attached hatchback ground vehicle (Ahmed body) due to a variable slant angle is accessed by this paper. The design includes a scaled Ahmed body as a reference ground vehicle and a rear wing with NACA 0018 profile. The computational studies are executed with Reynolds-averaged Navier–Stokes based k-epsilon turbulence model with nonequilibrium wall function. The vehicle's model is scaled to 75% of the actual model, and analyses are conducted with Reynolds number 2.7 × 106. After the study, it is observed that a 15 deg slant angle is the critical angle for the wing attached state in which the drag coefficient is maximum. After this angle, a sudden reduction of coefficients is observed, where 25 deg is critical for without wing condition. Besides this, the two counter-rotating horseshoe vortices in the separation bubble and side edge c-pillar vortices also behave differently due to the wing's presence. The turbulent kinetic energy variation and the variation in coefficients of surface pressure are also affected by the rear wing attachment. This paper will assist in finding the optimum slant angle for hatchback ground vehicles in the presence of a rear wing. Thus the study will help in increasing stability and control for hatchback ground vehicles.

Effect of size distribution of bubble groups on average volumetric heat transfer coefficient in the process of direct contact boiling heat transfer

A new metric, the average width of bubbles, which can be used to assessing the dynamic evolution of bubble groups in the process of direct contact boiling heat transfer, is proposed. The average width is defined as the average of distance from the bubble's edge to the centroid. The correlation model between the change of Betti number slope and the average volumetric heat transfer coefficient is established. The results show that the average volumetric heat transfer coefficient is reduced as the Betti number slope increased. Simultaneously, the size distribution modeling in the bubble groups growth direction is obtained and constants K are obtained by double-logarithm regression equation. The experimental data shows that the average width (y) of bubble groups is as a function of bubble growth direction stage (x) (y = ebxk). The ratio (η) based on average width normalized is proposed and an experimental optimization strategy to characterize the excellent heat transfer efficiency is established with η ≥ 4. The value of b has a significant positive impact on average volumetric heat transfer coefficient.

Spatial confinement effects of bubbles produced by laser ablation in liquids

The paper deals with the influence of the spatial confinement on the evolution of laser bubbles and shock waves by means of the shadowgraphy technique. Due to the constraint of walls, the bubble center shrinks faster than the bubble edge and collapses before the edge of the bubble, splitting into left and right small bubbles that continue to shrink down. The result validates that the Bjerknes force has little effect on bubble evolution at the expansion stage but a great influence on it at the collapse stage. We study the evolution of laser bubbles with different Al-plate intervals for displaying a gradual transition from constrained conditions to unconstrained conditions. In addition, we describe the dynamics of the first bubble at the expansion stage using the Rayleigh-Plesset equation. The pressure and temperature inside laser bubbles are calculated in the meanwhile.

An improved image procession algorithm for bubble flow characteristics in gas–liquid reactor

AbstractBubble size distributions in bubble‐liquid reactors are essentially determined by the multiphase turbulent hydrodynamic behaviors, which take greatly effects on the heat and mass transfer. An improved image procession algorithm, effectively combined for the first time, color space transformation, shadow detection and removal, and watershed segmentation to handle the raw pictures captured by high‐speed camera was proposed to characterize the distributions of bubble sizes. The hue–saturation–intensity color space approach to detect bubble edges and shadow regions for their intensity was set to as input value for processing background subtraction. Furthermore, the removal of bubble shadow was successfully performed by means of extracting shadow regions to attain the satisfied compensation images, a watershed transformation function was also developed to segment the overlapping bubbles into the individual bubble objects after shadow removed as well. Bubble parameters, that is, bubble areas and bubble perimeters are obtained details, and results showed that the maximum errors between calculated results and measurement without considering shadow effects are approximately 25%. Thus, it indicated that the effects of shadow are seriously, and it must be optimized for better prediction regarding the ability and the capacity of reactor design.

The Sunyaev–Zel’dovich Effect of Simulated Jet-inflated Bubbles in Clusters

Abstract Feedback by active galactic nuclei (AGNs) is essential for regulating the fast radiative cooling of low-entropy gas at the centers of galaxy clusters and for reducing star formation rates of central ellipticals. The details of self-regulation depend critically on the unknown contents of AGN-inflated bubbles. Observations of the Sunyaev–Zeldovich (SZ) signal of AGN bubbles provide us with the ability to directly measure the lobe electron pressure given a bubble morphology. Here we compute the SZ signal of jet-inflated bubbles in three-dimensional magnetohydrodynamical simulations of the galaxy cluster MS0735.6+7421 with the Arepo code, and compare our synthetic SZ results to inferences obtained with popular modeling approaches. We find that cutting out ellipsoidal bubbles from a double-beta pressure profile only matches the inner bubble edges in the simulations and fails to account for the emission of the shock-enhanced pressure cocoon outside the bubbles. This additional contribution significantly worsens the accuracy of the cut-out method for jets with small inclinations with respect to the line of sight. Also, the kinetic SZ effect of the bubbles, a previously neglected contribution, becomes relevant at these smaller inclinations due to entrainment and mixing of the intracluster medium with low-density jet material. Fortunately, the different signs of the kinetic SZ signal in opposite lobes allow this effect to be modeled. We present an approximate method to determine the jet inclination, which combines jet power and lifetime estimates, the stand-off distance between jet head and bow shock, and the kinetic SZ effect, thereby helping to correctly infer the bubble contents.

Probing the nanoscale origin of strain and doping in graphene-hBN heterostructures

We use confocal Raman microscopy and a recently proposed vector analysis scheme to investigate the nanoscale origin of strain and carrier concentration in exfoliated graphene-hexagonal boron nitride (hBN) heterostructures on silicon dioxide (SiO2). Two types of heterostructures are studied: graphene on SiO2 partially covered by hBN, and graphene fully encapsulated between two hBN flakes. We extend the vector analysis method to produce separated spatial maps of the strain and doping variation across the heterostructures. This allows us to visualise and directly quantify the much-speculated effect of the environment on carrier concentration in graphene. Moreover, we demonstrate that variations in strain and carrier concentration in graphene arise from nanoscale features of the heterostructures such as fractures, folds and bubbles trapped between layers. For bubbles in hBN-encapsulated graphene, hydrostatic strain is shown to be greatest at bubble centres, whereas the maximum carrier concentration is localised at bubble edges. Raman spectroscopy is shown to be a non-invasive tool for probing strain and doping in graphene, which could prove useful for engineering of two-dimensional devices.

How to measure the thickness of a lubrication film in a pancake bubble with a single snapshot?

In the in-line bright-field image of a pancake-like bubble, a ring-shaped zone of maximum intensity is visible, called the glare ring. It is due to multiple interactions of light with the bubble interface. In this study, we develop a method to measure the thickness of the lubrication film around a pancake-like bubble translating inside a microchannel, based on the location of this glare ring. By means of ray tracing, a correlation is proposed to relate the film thickness to the location of the glare ring with respect to the bubble edge and to the ratio of refractive indices of the inner and outer phases. This makes the method also applicable to inviscid pancake drops. Additionally, for static bubbles, the method can be used to measure the depth of a microchannel. For moving bubbles, provided the speed of the bubble is also measured, the method can be used to measure surface tension or viscosity. Finally, the method can also be extended to viscous drops, provided the shape function of the interface is adapted.

Adaptive fuzzy local ternary pattern for mineral flotation froth image edge detection

It is difficult to segment blur edge for mineral flotation froth image, an edge detection method based on fuzzy ternary pattern is proposed in this paper. On the basis of binary logic pattern, a kind of fuzzy logic state is introduced. According to the formation mechanism of bubble and the corresponding valley edge characteristics, the local fuzzy ternary pattern algorithm is adopted to describe uncertain intensity relationship between center and neighborhood pixel. Meanwhile, the bidirectional grey scale difference is fuzzed to describe the relationship between edge and non-edge. On this basis, the memberships of neighborhood intensity and bidirectional intensity difference are combined, and the bubble edge membership matrix is built. Thus, the candidate boundary pixels are determined by the joint membership. According to candidate pixels characterized, some non-boundary pixels are removed, and bubble edges are reserved. The experimental results show that the proposed method can effectively detect the bubble edges.