S Pushpavanam

Liquid-liquid slug flow regime is characterised by the presence of strong internal circulation which enhances mass transfer. In this work, we develop a mathematical model for conjugate mass transfer in the liquid-liquid slug flow regime in a microchannel. A Lagrangian approach is adopted where the behaviour of a unit cell in the channel is analysed in a moving reference frame. The system is analysed for two cases when (I) the slug is in direct contact with the channel wall and (II) there is a thin film of continuous fluid surrounding the slug. A novel contribution of this work is the extension of the stream function formulation to determine the flow structures in multiply connected domains made up of two liquids in a pressure driven flow. The primary objective of the work is to understand the influence of the flow patterns (which is determined by fluid properties and operating conditions) on the mass transfer of a solute from the slug to the continuous phase. Towards this, the species transport equation is solved numerically using the velocity field obtained. The reliability of the model developed is validated with experiments reported in the literature. This study gives us insights on the influence of the film on the hydrodynamics and its contribution to mass transfer. A key finding of this study is that the rate of mass transfer can be enhanced if the continuous phase has higher viscosity than the slug phase.

The effect of a soluble surfactant on the linear stability of layered two-phase Poiseuille flows through soft-gel-coated parallel walls is studied in this paper. The focus is on determining the effect of the elastohydrodynamic coupling between the fluids and the soft-gel layers on the various flow instabilities. The fluids are assumed Newtonian and incompressible, while the soft gels are modeled as linear viscoelastic solids. The effect of a soluble surfactant on the different instabilities is specifically investigated. The soft-gel-coated plates are maintained at two different solute concentrations. The dynamics of the soluble surfactant in the fluids is captured using a species transport equation. A linear stability analysis is carried out to identify different instabilities in the system. The linearized governing equations are solved numerically using a Chebyshev spectral Collocation technique. The effect of deformability of the soft gels on three distinct instability modes, (a) a liquid-liquid long-wave mode, (b) a liquid-liquid short-wave mode, and (c) a liquid-liquid Marangoni short-wave mode, is analyzed. An analytical expression for the growth rate is obtained in the long-wave length limit using an asymptotic analysis. From the long-wave analysis a stability map is obtained, in which dominant effects in different regions are identified. The Marangoni stresses can either stabilize or destabilize the interfacial instability depending on the direction of mass transfer. They have a predominantly stabilizing effect on the interfacial instability when the mass transfer is from the more viscous broader fluid to the less viscous thinner fluid. Placing a gel closer to the more viscous fluid has a stabilizing effect on this instability. The Marangoni stresses and soft-gel layers can have opposing effects on the stability of the long-wave mode. The dominant of these two opposing effects is determined by the prevailing parameters. Insights into the dominant physical causes of different instabilities are presented.

We numerically investigate the hydrodynamics of a two-dimensional compound drop in a plane Poiseuille flow under Stokes regime. A neutrally buoyant, initially concentric compound drop is released into a fully developed flow, where it migrates to its equilibrium position. Based on the results, we find that the core-shell interaction affects the dynamics of both the core and the compound drop. During the initial transient period, the core revolves about the center of the compound drop due to the internal circulation inside the shell. At equilibrium, depending upon the nature of the flow field inside the shell, we identify two distinct core behaviors: stable state and limit-cycle state. In the stable state, the core stops revolving and moves outward very slowly. The core in the limit-cycle state continues to revolve in a nearly fixed orbit with no further inward motion. The presence of the core affects both deformation and migration dynamics of the compound drop. A comparison with the simple drop reveals that the core enhances the deformation of the compound drop. The outward moving core in the stable state pushes the compound drop toward the walls, while the revolving core in the limit-cycle state causes the compound drop to oscillate at its equilibrium position. The migration of the compound drop also affects the eccentricity of the core significantly. From the parametric study, we find that the core affects the compound drop dynamics only at intermediate sizes, and an increase in any parameter sufficiently causes a transition from the limit-cycle state to the stable state.

Bubble induced liquid circulation is important in applications such as bubble columns and air-lift reactors. In this work, we describe an experimental and numerical investigation of liquid circulation induced by a bubble plume in a tank partitioned by a baffle. The baffle divides the tank into two compartments. Liquid can flow from one compartment to the other through openings at the top and the bottom of the baffle. Gas (air) was injected in the riser section in the form of bubbles at one corner of the tank. The temporal and spatial variation of velocity field in the liquid as a function of the gas flow rate was measured using particle image velocimetry (PIV). At a constant gas flow rate, the liquid flow field is unsteady due to the interaction with the bubbles. The time scales associated with the velocity-time series and the bubble plume thickness variation were calculated. The time averaged-velocity field was used to quantify the variation of the liquid circulation rate with gas flow rate. The turbulence in the liquid was measured in terms of turbulent intensities. These were calculated from the experimental data and were observed to be less than 3 cm/s. A 2-d Euler-Euler two-fluid model with buoyancy and drag as the interaction terms was used to simulate the flow. The parameters chosen for the simulations were selected from literature. It is shown that inclusion of turbulence model such as k - ε{lunate} is necessary to capture the overall flow behavior. Good agreement was observed between experimentally obtained velocity profiles and the recirculation rates with the simulation results. © 2007 Elsevier Ltd. All rights reserved.

The time-dependent fluid flow in a square cavity was studied using model fluids of glyceml-water solution at different frequencies and amplitudes of motion of the top plate. The range of Reynolds numbers in our investigation varied from 5 to 3700. The experiments were carried out in a square cavity with a periodically driven lid, and planar velocity measurements were obtained using particle image velocimetry. The flow was driven by moving the top surface of the cavity in a simple harmonic motion. The aspect ratio, defined as the ratio of cavity length to the cavity height, is unity. The ratio of cavity spanwise width to the length of the cavity is 0.2. The temporal variation of velocity at fixed locations in the cavity exhibits a periodic variation. The basic frequency of the fluid motion at a point in the flow domain was observed to be the same as that of plate motion for low Reynolds number Re. However, existence of dominant secondary frequencies was observed along the central vertical plane. The velocity variation as a function of time at a fixed position and the velocity profiles along horizontal and vertical planes are also quantitatively described. These were compared to computational fluid dynamics (CFD) simulations based on the finite volume technique. Comprehensive details of the flow as a function of Reynolds number are analyzed. The evolution of secondary vortices at different plate positions as a function of Reynolds number is also presented. The planar velocity measurements acquired are indicative of the flow behavior in a periodically driven cavity with a narrow span width even at high Re. At very low Re, the flow throughout the periodically driven cavity qualitatively resembles the classical steady lid-driven cavity flow. At high Re, the entire cavity is occupied with multiple vortices. The qualitative features of the bulk flow observed are valid even for cavities with infinite span width. Copyright © 2006 by ASME.

Abstract: Chemically active Janus particles generate tangential concentration gradients along their surface for self-propulsion. Although this is well studied in unbounded domains, the analysis in biologically relevant environments such as confinement is scarce. In this work, we study the motion of a Janus sphere in weak confinement. The particle is placed at an arbitrary location, with arbitrary orientation between the two walls. Using the method of reflections, we study the effect of confining planar boundaries on the phoretic and hydrodynamic interactions, and their consequence on the Janus particle dynamics. The dynamical trajectories are analyzed using phase diagrams for different surface coverage of activity and solute-particle interactions. In addition to near wall states such as ‘sliding’ and ‘hovering’, we demonstrate that accounting for two planar boundaries reveals two new states: channel-spanning oscillations and damped oscillations around the centerline, which were characterized as ‘scattering’ or ‘reflection’ by earlier analyses on single wall interactions. Using phase-diagrams, we highlight the differences in inert-facing and active-facing Janus particles. We also compare the dynamics of Janus particles with squirmers for contrasting the chemical interactions with hydrodynamic effects. Insights from the current work suggest that biological and artificial swimmers sense their surroundings through long-ranged interactions, that can be modified by altering the surface properties. Graphic abstract: [Figure not available: see fulltext.].

Microfluidic paper-based analytical devices (µPADs) have provided a breakthrough in portable and low-cost point-of-care diagnostics. Despite their significant scope, the complexity of fabrication and reliance on expensive and sophisticated tools, have limited their outreach and possibility of commercialization. Herein, we report for the first time, a facile method to fabricate µPADs using a commonly available laser printer which drastically reduces the cost and complexity of fabrication. Toner ink is used to pattern the µPADs by printing, without modifying any factory configuration of the laser printer. Hydrophobic barriers are created by heating the patterned paper which melts the toner ink, facilitating its wicking into the cross-section of the substrate. Further, we demonstrate the utilization of the fabricated device by performing two assays. The proposed technique provides a versatile platform for rapid prototyping of µPADs with significant prospect in both developed and resource constrained region.

During the anodic dead-end mode operation of fuel cells, the inert gases (nitrogen and water) present in the cathode side gas channel permeate to the anode side and accumulate in the anode gas channel. The inert gas accumulation in the anode decreases the fuel cell performance by impeding the access of hydrogen to the catalyst. The performance of fuel cell under potentiostatic dead-end mode operation is shown to have three distinct regions viz. time lag region, transient current region and a steady state current region. A current distribution measurement setup is used to capture the evolution of the current distribution as a function of time and space. Co- and counter-flow operations of dead-end mode confirm the propagation of inert gas from the dead-end of anode channel to the inlet of anode. Experiments with different oxidants, oxygen and air, under dead-end mode confirm that nitrogen which permeates from cathode to anode causes the performance drop of the fuel cell. For different starting current densities of 0.15 A cm-2, 0.3 A cm-2 and 0.6 A cm-2 the inert gas occupies 35%, 45% and 57%, respectively of anode channel volume at the end of 60 min of dead-end mode operation. © 2011 Elsevier B.V. All rights reserved.

In this work, we investigate experimentally as well as numerically a drainage displacement system; i.e., a non-wetting fluid displacing a wetting fluid in a porous medium. Experiments were carried out in a horizontal rectangular channel packed with a monolayer of glass beads. The displacement of a higher viscosity wetting fluid (silicone oil) by a lower viscosity non-wetting fluid (air) is studied. Similarly, the displacement of a lower viscosity wetting fluid (silicone oil) by a higher viscosity non-wetting fluid (glycerol) is also studied. Flow structures, such as viscous fingering and stable displacement, were obtained. The behavior of the flow in the experiments was simulated using a pore network model. The model consists of a network of tubes of equal lengths inclined at 45°. The radius of the tubes is assumed to follow a random distribution to ensure a realistic representation of a porous medium. The pressure distribution across the network is obtained by assuming laminar flow in each tube. The Hagen-Poiseuille equation is used after including the effect of capillary pressure to determine the flow velocity in each tube. The displacement of the interface for each time step is restricted to 2.5-5.0% of the tube length and the maximum velocity in the network is used to calculate this time interval. The movement of the interface inside the tube is calculated using a second-order Runge-Kutta method. Once the interface reaches a node, the volume of the fluid entering the neighboring tubes is determined by the pressure drop across them. We have varied the capillary number, Ca (μv/σA), and viscosity ratio, M, and have obtained two different flow regimes, viscous fingering and stable displacement. The residual amount of defending fluid present in the network model is calculated for the two regimes of drainage displacements. It is found that when stable displacement occurs, the system has significantly less amount of defending fluid present for the same duration of time as compared with the case when viscous fingering is exhibited. The fronts of the invading fluid during viscous fingering at different capillary numbers are self-similar with a fractal dimension of 1.3 that matches with the experimental results. © 2011 by Begell House, Inc.