Volume of fluid method

About: Volume of fluid method is a research topic. Over the lifetime, 5338 publications have been published within this topic receiving 116760 citations.

Experimental and numerical investigation of capillary flow in SU8 and PDMS microchannels with integrated pillars

Abstract: Microfluidic channels with integrated pillars are fabricated on SU8 and PDMS substrates to understand the capillary flow. Microscope in conjunction with high-speed camera is used to capture the meniscus front movement through these channels for ethanol and isopropyl alcohol, respectively. In parallel, numerical simulations are conducted, using volume of fluid method, to predict the capillary flow through the microchannels with different pillar diameter to height ratio, ranging from 2.19 to 8.75 and pillar diameter to pitch ratio, ranging from 1.44 to 2.6. The pillar size (diameter, pitch and height) and the physical properties of the fluid (surface tension and viscosity) are found to have significant influence on the capillary phenomena in the microchannel. The meniscus displacement is non-uniform due to the presence of pillars and the non-uniformity in meniscus displacement is observed to increase with decrease in pitch to diameter ratio. The surface area to volume ratio is observed to play major roles in the velocity of the capillary meniscus of the devices. The filling speed is observed to change more dramatically under different pillar heights upto 120 μm and the change is slow with further increase in the pillar height. The details pertaining to the fluid distribution (meniscus front shapes) are obtained from the numerical results as well as from experiments. Numerical predictions for meniscus front shapes agree well with the experimental observations for both SU8 and PDMS microchannels. It is observed that the filling time obtained experimentally matches very well with the simulated filling time. The presence of pillars creates uniform meniscus front in the microchannel for both ethanol and isopropyl alcohol. Generalized plots in terms of dimensionless variables are also presented to predict the performance parameters for the design of these microfluidic devices. The flow is observed to have a very low Capillary number, which signifies the relative importance of surface tension to viscous effects in the present study.

Modeling metamorphic fluid flow with reaction-compaction permeability feedbacks

Abstract: Existing models of metasomatic flow do not allow for the effect that reaction has on the flow patterns. Instead, it is assumed that the volatiles produced are negligible in volume compared to those infiltrated and that reaction does not modify permeability. This is clearly unlikely to be true for infiltration-driven decarbonation reactions. The rates of porosity creation by reaction and porosity loss by creep have been calculated for a representative volume of calcite –quartz-wollastonite marble, and it is found that, even for a weak calcite matrix, the rate of porosity generation by reaction is likely to outstrip the collapse of porosity, as long as the system is out of equilibrium. We have applied a self-consistent 1D finite-difference model to the reaction of calcite + quartz to wollastonite in a 10m thick marble, in response to influx of H2O rich fluid, with fixed boundary conditions. The model allows us to evaluate the effect of reaction on the porosity structure and fluid pressure variation across the layer, from which local Darcy fluxes can be evaluated. The progress of reaction that we model is constrained by hydrological considerations, with the requisite parameters recalculated as reaction progresses, assuming creep compaction of rock under the stress difference between lithostatic and fluid pressures. We fnd that the volume of fluid realised by decarbonation, driven by influx of H20, is sufficient to create a back-flow, so that further advancement of the reaction front is only possible as a result of diffusion of water against the Darcy flux. The effect of creep driven by differences between fluid pressure and lithostatic pressure is to reduce the permeability of the layer and especially reduce the secondary porosity developed in the zone at and behind the advancing reaction front. We predict that in a 3D situation, the porous zone of reacted marble becomes a conduit for layer-parallel flow, and the secondary porosity is infilled by calc-silicate minerals due to silica metasomatism.

Transition from spherical cap to toroidal bubbles

Abstract: Large gas bubbles rising under the effect of buoyancy are known to adopt either a spherical cap shape or to undergo a topological transition after which they become toroidal. We carry out an axisymmetric numerical investigation of the evolution of such large bubbles in the presence of both capillary and viscous effects. The numerical approach is of the volume of fluid type (it solves the Navier-Stokes equations on a fixed grid and transports the local volume fraction of one of the fluids), but does not involve any explicit reconstruction of the interface. The transition from spherical cap to toroidal bubbles is studied in the parameter space built on the Bond (Bo) and Archimedes (Ar) numbers, which compare the strength of inertial effects to that of capillary and viscous effects, respectively. Preliminary tests show that the position of this transition is very sensitive to the grid resolution; these tests are used to select grid characteristics that yield grid-independent results. Two markedly different transition scenarios, corresponding to the limit of large Ar and large Bo, respectively, are then identified. In the first case, the front of the bubble is pierced by an upward jet coming from the rear of the bubble. In contrast, in the limit of large Bo, a downward jet develops at the front part and pierces the rear of the bubble, unless viscous effects are sufficient to stabilize the front. We also determine the position of the transition for intermediate values of Bo and Ar and discuss the connection between present axisymmetric results and experimental situations in which the bubble is followed by a turbulent wake. We finally examine a puzzling feature of these large bubbles which is that, given an initial gas volume, the final bubble topology appears to depend dramatically on the initial conditions. Indeed, we find that initially oblate bubbles may result in stable spherical cap bubbles for values of Bo and Ar well beyond those for which initially spherical bubbles of similar volume undergo the topological transition. This remarkable influence of the initial shape is shown to be due to the influence of the oblateness on both the bubble acceleration and the hydrostatic pressure difference between the two bubble poles.