3 results
Now showing1 - 3 of 3
- Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , A lagrangian model for the prediction of bubble breakupIn many applications in the chemical or biochemical industry the mass transfer of a species from one phase into another has a huge influence on the whole process. Often, the relevant species are available as gas and, thus brought into the observed system as gaseous bubbles. The transport of mass from those bubbles into the liquid takes place at the bubbles’ surfaces. Bubbles can break up, resulting in two or more daughter bubbles, increasing the total surface area and coalesce with each other, merging two or more bubbles into one, reducing the total surface area. Further impact on the mass transfer lies in the sizes of the boundary layers surrounding the bubbles and the local concentration differences, resulting in a difference in the chemical potential, being the main driver of the transfer itself. In this work, the focus lies on the derivation, implementation, and validation of the newly proposed model to describe the breakup of gaseous bubbles. For this purpose, a trajectory-based attempt is used while each bubble is modelled individually as Kelvin-Voigt element. Further, the bubbles are assumed to behave like rotational ellipsoids in the current state of the model. The interfacial tension is mirrored by a theoretical spring, inducing a tension on each observed bubble forcing it into its equilibrium, the spherical state. The gaseous phase’s viscosity is depicted by a theoretical damper, acting against the current direction of the bubble’s further deformation or restoration. The tension which drives the deformation is retrieved from a Lagrangian analysis, quantifying the local occurring stretch implied by the surrounding liquid. The deformed bubble is then destabilised in comparison to the spherical state. For the description of the breakup the force induced by interfacial tension, lowered by the deformation, is compared to the inertial forces acting on the bubble. These inertial forces are quantified by the drag force which acts on the bubble. In the Trajectory-Based Breakup Model, the drag force is approximated by the buoyancy force. This description has a potentially severe disadvantage. There is one critical value for the state of deformation for a given bubble volume, regardless the flow conditions of the liquid. To account for this, a more precise formulation of the drag force is derived. Both model variations are tested by carrying out Lattice Boltzmann simulations. The mass transfer is modelled based on the retrieved bubbles sizes coupled with either the penetration theory or a Sherwood correlation. The results are compared to experimental data for validation.Publicationtype: Doctoral Thesis76 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Direct numerical simulation of bubble collision, bounce and coalescence in bubble-induced turbulence(2024-02-05); ; ; A direct numerical simulation study has been carried out to better understand the dynamics of deformable bubbles in bubble-induced turbulence. A mathematical model for bubble collision, bounce and coalescence is developed based on the Volume-of-Fluid (VoF) method. A surface-tension-holding film thickness δc is introduced to determine whether the bubbles will coalesce or bounce after collision. Bubbly flows with the gas volume fractions αg=5.6% (low), 11.2% (medium) and 22.1% (high) are considered in the study. The DNS results show that the dissipation of turbulent kinetic energy in the cases under consideration takes place when the wavenumber κ is smaller than twice the wavenumber for the bubble diameter 2κd. The bubble-induced turbulence is anisotropic at the length scales close to the bubble diameter due to the elongated turbulent structures in the bubble-rising direction. The bubbles collide intensively with each other when they are in a cluster surrounded by the liquid with low pressure. Parallel clustering of bubbles is found when the gas volume fraction has a low or medium value. Three-dimensional clustering is found when the bubbles are densely populated. Two different mechanisms of bubble collision have been identified from the DNS results, termed parallel collision and vertical collision. Parallel collision is often observed when bubbles are sparsely populated. In a parallel collision, the relative velocity of the bubbles slows down as two bubbles approach each other due to the jet flow between them, while the relative velocity increases sharply after the bounce. In a vertical collision, by contrast, the relative velocity of bubbles accelerates as two bubbles approach each other, while it slows down during the collision. Vertical collisions occur when the bubbles are more densely populated. The numerical results also show the significant effects of δc on bubble coalescence.Publicationtype: Journal Article18 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Effects of bubble-induced turbulence on interfacial species transport: A direct numerical simulation study(Elsevier, 2023-09-05); ; ; A direct numerical simulation (DNS) study is carried out to understand the effects of bubble-induced turbulence on the interfacial species transport. A volume-of-fluid (VOF) method is used in the DNS to simulate the multiphase flow in a box containing one or two deformable bubbles. The bubbly flows with Schmidt numbers 1≤Sc≤64, Reynolds numbers 100≤ReB≤750, Eötvös number Eo=1.21 and gas volume fractions 2.8%≤αg≤22.1% have been investigated. The DNS results indicate that the one-dimensional energy spectra evolve as the power −3 of the wavenumber at large scales. At a high Schmidt number, an inertial subrange characterized by the -5/3 slope can be found in the power spectra of species concentration. The power spectra of species concentration for different Schmidt numbers become close to each other when the wavenumber is scaled with Sc0.5. The bubble-induced turbulence enhances the oscillation of the transient Sherwood number, however, it has marginal effects on the time-averaged value. By contrast, the time-averaged Sherwood number for a fixed Peclet number increases with an increase of αg, indicating that convection plays a more important role in species transport when the bubbles are more densely populated. A possible reason is that, the neighboring bubbles enhance the spatial fluctuations of the velocity, which favor the interfacial species transfer. The energy spectrum confirms that an increase of αg leads to stronger spatial fluctuations of the vertical velocity component at low wavenumbers.Publicationtype: Journal Article38