Bulletin of the American Physical Society
2005 58th Annual Meeting of the Division of Fluid Dynamics
Sunday–Tuesday, November 20–22, 2005; Chicago, IL
Session AE: Multiphase Bubbly Flows I |
Hide Abstracts |
Chair: Said Elghobashi, University of California, Irvine Room: Hilton Chicago Continental B |
Sunday, November 20, 2005 8:00AM - 8:13AM |
AE.00001: Effects of microbubbles on the Taylor-Green vortex flow A. Ferrante, S. Elghobashi Numerical simulations of the Taylor-Green vortex (TGV) flow laden with microbubbles were performed to study the effects of microbubbles on a simple vortical flow using the two-fluid approach. The study was motivated by our DNS results of a spatially developing turbulent boundary layer laden with microbubbles [J. Fluid Mech. 503 (2004)] which showed that the presence of bubbles results in a {\em local positive divergence of the fluid velocity}, $\nabla \cdot \mathbf{U}$. This velocity divergence displaces the near-wall quasi-streamwise vortical structures away from the wall, thus reducing the skin friction. In the present study, the continuity and momentum equations of both phases (fluid and bubbles) were numerically solved in a cubical domain. The results for Stokes number equal to 0.25 and bubbles volume fraction of $1\%$ show that the magnitude of the vorticity at the center of the vortex decays faster than that of the single-phase flow. After 20 turnover times of the initial vortex, the magnitude of the vorticity at the center of the vortex becomes 30\% smaller than that of the single-phase flow. Analysis of the vorticity equation shows that the {\em local positive velocity divergence of the fluid velocity}, $\nabla \cdot \mathbf{U}$, created in the vortex core by the clustering of the bubbles, is responsible for the vorticity decay. Results for different Stokes numbers and bubbles volume fractions will be presented. [Preview Abstract] |
Sunday, November 20, 2005 8:13AM - 8:26AM |
AE.00002: Laminar bubbly flow in a vertical channel Jiacai Lu, Souvik Biswas, Gretar Tryggvason Direct numerical simulations are used to examine the buoyant rise of many nearly spherical bubbles in laminar flows in vertical channels. The lift force on spherical bubbles leads to a very simple flow structure in terms of the void fraction distribution and the average liquid velocity. The numerical results show that at steady state the number density of bubbles in the center of the channel is always such that the fluid mixture there is in hydrostatic equilibrium and the velocity is uniform. For upflow, excess bubbles are pushed to the walls, forming a bubble rich layer, one bubble diameter thick. For downflow, bubbles are drawn into the channel core, leading to a wall layer with no bubbles, of a thickness determined by the pressure gradient and the average void fraction. For the downflow, the void fraction profile and the velocity profile can be predicted analytically, but for upflow the velocity increase across the wall-layer must be obtained from the simulations. The behaviour of the bubbles in the middle of the channel, including the slip velocity and their velocity fluctuations, is well predicted by results for homogeneous flows in fully periodic domains. [Preview Abstract] |
Sunday, November 20, 2005 8:26AM - 8:39AM |
AE.00003: Influence of surfactant conditions on the structure of an upward bubbly channel flow Toshiyuki Ogasawara, Shu Takagi, Yoichiro Matsumoto We investigated an upward bubbly channel flow and the effects of surfactant on its flow structure experimentally. 3-Pentanol and Triton X-100 are used as surfactants. By the addition of small amount of surfactant, bubble coalescences are prevented and mono-dispersed 1mm spherical bubbles are obtained. Under all of our experimental conditions, the added surfactants do not influence the single-phase turbulence. On the other hand, small amount of surfactant drastically changes the whole flow structure of bubbly flow. On the low concentration of 3-Pentanol (21-63ppm), bubbles strongly migrate towards the wall and these highly accumulated bubbles on the wall form crescent-like shaped horizontal bubble clusters of 10-40mm length. However, in 3-Pentanol solution of higher concentration ($\sim $168ppm) or in the 2ppm Triton X-100 solution, the tendency of the lateral migration of bubbles is weaken and the bubbles are distributed uniformly in the channel. In the surfactant solution, the slip velocity on the bubble surface retards and the bubble rising velocity decreases (Marangoni effect). The change of boundary condition on the bubble surface affects not only drag force but shear-induced lift force. It is indicated that this change of shear-induced lift force greatly relates to the lateral migration of bubbles and the disaggregation of the bubble clusters. We also measured the turbulent properties using LDV and discuss the flow structure. [Preview Abstract] |
Sunday, November 20, 2005 8:39AM - 8:52AM |
AE.00004: Bubbly wall-layers in a vertical channel Souvik Biswas, Jiacai Lu, Gretar Tryggvason Direct numerical simulations of nearly spherical bubbles rising in a laminar flow in vertical channels have shown that for upflow the bubbles are pushed to the walls, until the fluid mixture in the center of the channel is in hydrostatic equilibrium. The excess bubbles hug the channel wall, forming a wall-layer, one bubble diameter thick. The upward velocity of the core flow depends entirely on the velocity increase across the wall layer. Here we examine how the bubbles in the wall layer rise and how their rise velocity, as well as the velocity in the center of the channel, depends on the governing parameters of the flow. The study is done using direct numerical simulations where the flow around the bubbles is fully resolved and the uniform flow outside the wall layer is generated by a properly adjusted body force. The behavior of the flow is studied for a range of parameters using a regular periodic array and the results then compared with results from simulations of freely evolving and interacting bubbles for one case, as well as with results of simulations of the full channel. The average properties of the flow in the wall layer are examined and compared with a simple two-fluid model. [Preview Abstract] |
Sunday, November 20, 2005 8:52AM - 9:05AM |
AE.00005: Near Wall Bubble Transport in a Forced Turbulent Boundary Layer David Jeon, Mory Gharib Transport of bubbles in turbulent boundary layers remains an area of active research. One of the areas of recent interest is the use of bubbles in skin friction drag reduction. However, for drag reduction to be effective, it seems that bubbles need to be kept in the near wall region, where wall shear stress derives from. Simulating the conditions meaningful to full scale vessels is very difficult in the laboratory due to scaling issues. Towards that end, we have used the idea of forced turbulence to simulate the near wall region. This allows us to inject bubbles into what is effectively the sub-layer, letting us explore bubble transport very close to the wall. We used the hydrogen wire technique to generate bubbles through electrolysis of water. The generating wire was placed at various heights above the wall to measure how transport is affected by injection location. Results indicate that injection at the wall may not be optimal with regards to keeping the bubbles near the wall. The authors would like to thank the Office of Naval Research for their support under Grant No. N00014-00-1-0110. [Preview Abstract] |
Sunday, November 20, 2005 9:05AM - 9:18AM |
AE.00006: Effect of bubble size on micro-bubble drag reduction Xiaochun Shen, Steven Ceccio, Marc Perlin The effect of bubble size on micro-bubble drag reduction was investigated experimentally in a high-speed turbulent channel flow of water. A variety of near-wall injection techniques were used to create a bubbly turbulent boundary layer. The resulting wall friction force was measured directly by a floating element force balance. The bubble size was determined from photographic imaging. Using compressed nitrogen to force flow through a slot injector located in the plate beneath the boundary layer of the tunnel test section, a surfactant solution (Triton X-100, 19ppm) and salt water solution (35ppt) generated bubbles of average size between $\sim $500 microns and $\sim $200 microns and $\sim $100 microns, respectively (40 $< \quad d^{+} \quad <$ 200). In addition hollow spherical glass beads ($\sim $75 microns ($d^{+}$ = 30) and specific gravity 0.18) and previously prepared lipid stabilized gas bubbles of $\sim $ 30 micron ($d^{+}$ =12) were injected. The results indicate that the drag reduction is related strongly to the injected gas volume flux and the static pressure in the boundary layer. Changing bubble size had essentially no influence on the measured friction drag, suggesting that friction drag is not a strong function of bubble size. [Sponsored by the Office of Naval Research.] [Preview Abstract] |
Sunday, November 20, 2005 9:18AM - 9:31AM |
AE.00007: A numerical study on the effect of the bubble diameter on the mass transfer in bubbly plumes Xiaobo Gong, Shu Takagi, Huaxiong Huang, Yoichiro Matsumoto A numerical simulation has been conducted for studying the effect of bubble diameter on the mass transfer efficiency and the concentration distribution of the dissolved gas in a bubbly plume. The numerical method for describing the bubbly plume with mass transfer was developed in an Euler-Lagrange way. The Navier-Stokes equation was adopted for the movement of the liquid phase. The motion of bubbles was tracked individually. The interaction between the liquid and bubbles were considered with a two-way coupling method. The model for the correlation of the dissolution and diffusion of the gas and the translational motions of bubbles with mass loss was introduced. The oxygen bubble plume in a quasi two dimensional rectangular water tank was simulated and studied. The numerical results show that the mass transfer efficiency non-linearly deceases with the increase of bubble diameter. Optimal bubble diameter exists for the mass transfer with a given tank size. The bubble diameter distribution of a certain range does not clearly affect the mean mass transfer efficiency. However, the mixing of different sizes of bubbles improves the uniformity of the concentration distribution in the flow field. [Preview Abstract] |
Sunday, November 20, 2005 9:31AM - 9:44AM |
AE.00008: Droplet Dispersion and Vapor Mixing by Fine Scale Turbulence Makoto Sato, Mamoru Tanahashi, Toshio Miyauchi Direct numerical simulations of evaporating droplet in homogeneous isotropic turbulence have been conducted to clarify the relationship between droplet dispersion and coherent fine scale eddies in turbulence. The dispersions of $10^6$ droplets are analyzed for several initial droplet Stokes numbers. The Stokes number that causes specific distribution of droplets is closely related to the time scales of coherent fine scale eddies. The number density of droplet with particular Stokes number is low near the center of the coherent fine scale eddies and shows the maxima near the radius of the eddy. Comparisons between non-evaporating and evaporating cases suggest that the diameter change due to evaporation strongly affect the droplet dispersion around the fine scale eddies. The range of the specific Stokes number for evaporating droplets is wider than that of non-evaporating droplets, which causes particular vapor concentration fluctuation in turbulence. [Preview Abstract] |
Sunday, November 20, 2005 9:44AM - 9:57AM |
AE.00009: Single-pressure and multi-pressure models for multiphase flows Duan Zhang, Brian VanderHeyden, Qisu Zou, Nely Padial-Collins In many multiphase flow models, only one pressure appears in the momentum equations for all the phases. For disperse multiphase flows, this pressure is usually chosen to be the pressure of the continuous phase. The pressure of the disperse phase is simply related to the pressure of the continuous phase by, for instances, adding surface tension. We refer to these types of multiphase flow models as single pressure models. While these types of models have been proven successful in many computations of disperse two-phase flows, they encounters conceptual difficulties when applied to continuous multiphase flows. For instance, the single pressure model cannot be used to study the tension break of a sponge with interconnected pores because the air in the pores of the sponge can never go into tension, while the sponge material can never break under compression. To avoid these conceptual difficulties, a multi-pressure model is introduced by analyzing assumptions related to the single pressure model. It is found that the applicability of the single pressure model relies on the validity of the small particle approximation. When the length scale of the ``disperse phase'' is comparable to the length scale of the flow, the small particle approximation fails and so does the single pressure model. Physical and numerical aspects of the multi-pressure model are discussed. Examples calculated using the multi-pressure model are presented. *Work carried out under the auspices of U.S. DOE. [Preview Abstract] |
Sunday, November 20, 2005 9:57AM - 10:10AM |
AE.00010: Micromorphic theory of multiphase immiscible mixtures Weiming Li, Samuel Paolucci We consider a general multiphase immiscible mixture whose individual components are separated by infinitesimally thin interfaces. General average balance equations for the different phases as well as for the overall mixture are derived by using a systematic spatial averaging procedure. To account for local micro-motions and micro-deformations, we model the mixture using micromorphic theory. A minimal determinate theory is obtained by taking an appropriate number of moments of the microelement balance equations for mass, momentum and energy, together with the production of entropy inequality. The resulting average balance equations include equations for microinertia and microspin tensors. These equations, together with appropriate constitutive equations consistent with the entropy inequality, enable the modeling of immiscible multiphase materials where internal field quantities, such as the volume fraction of different phases, are important. To demonstrate the generality of the results, we apply it to a bubbly fluid. We show that the equations for microspin and microinertia, under a number of simplifying assumptions, combine to yield a general form of the Rayleigh-Plesset equation. Such an equation, in addition to accounting for the local average bubble microstretch (expansion or contration), also accounts for the local average bubble microrotation. Moreover, it contains higher-order microstructural statistics which in this minimal theory can be modeled by a constitutive approximation. [Preview Abstract] |
Sunday, November 20, 2005 10:10AM - 10:23AM |
AE.00011: Study on Fins' Effect of Boiling Flow in Millimeter Channel Heat Exchanger Satoshi Watanabe, Koji Okamoto, Yasuyuki Imai Recently, a lot of researches about compact heat exchangers with mini-channels have been carried out with the hope of obtaining a high-efficiency heat transfer, due to the higher ratio of surface area than existing heat exchangers. However, there are many uncertain phenomena in fields such as boiling flow in mini-channels. Thus, in order to understand the boiling flow in mini-channels to design high-efficiency heat exchangers, this work focused on the visualization measurement of boiling flow in a millimeter channel. A transparent acrylic channel (heat exchanger form), high-speed camera (2000 fps at 1024 x 1024 pixels), and halogen lamp (backup light) were used as the visualization system. The channel's depth is 2 mm, width is 30 mm, and length is 400 mm. In preparation for commercial use, two types of channels were experimented on: a fins type and a normal slit type (without fins). The fins are circular cylindrical obstacles (diameter is 5 mm) to promote heat transfer, set in a triangular array (distance between each center point is 10 mm). Especially in this work, boiling flow and heat transfer promotion in the millimeter channel heat exchanger with fins was evaluated using a high-speed camera. [Preview Abstract] |
Sunday, November 20, 2005 10:23AM - 10:36AM |
AE.00012: Mechanism of air entrainment in deep-water hydraulic jump Alberto Aliseda, Javier Rodriguez-Rodriguez, Juan Lasheras Based on experimental observations, we propose a new mechanism for air entrainment in a deep-water hydraulic jump. It is shown that in a region close to the toe of the jump, the coherent flow structures resemble those in two-dimensional mixing layers widely described in the literature. In the entrainment region, properly defined local Froude and Weber numbers are very high, thus buoyancy and surface tension effects can be neglected as a first approximation. We show that air entrainment is produced by the engulfment of big cavities of ``fresh fluid'' by coherent structures. By extending available models of the mixing layer to the large density ratio present in a hydraulic jump, the air entrainment in this type of flows can be estimated. PIV measurements of the velocity field in an attatched deep-water hydraulic jump as well as high-speed visualizations are presented to support the proposed mechanism. The structure of this mixing layer collapses at a certain distance downstream, as the local Froude number decreases and buoyancy effects are no longer negligible. At this point, where the majority of air entrainment has occurred, this mechanism no longer contributes to the overal increase in void fraction. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700