Modeling of low-capillary number segmented flows in microchannels using OpenFOAM

Abstract

Abstract. Modeling of low-Capillary number segmented flows in microchannels is important for the design of microfluidic devices. We present numerical validations of microfluidic flow simulations using the volume-of-fluid (VOF) method as implemented in OpenFOAM. Two benchmark cases were investigated to ensure the reliability of OpenFOAM in modeling complex physical phenomena in microfluidics, viz. 1) the steady motion of bubbles in capillaries, and 2) the formation of bubbles in T-junctions. We found that it is crucial to reduce spurious currents and to apply local grid refinement to capture the relevant flow physics. With these, we obtain good agreement between our numerical simulations and previously published theoretical and experimental data.

Modeling the motion of a Taylor bubble in a microchannel through a shear-thinning fluid

“…We solve a structured grid of nx?ny squared elements. Similar works [41,42] show that the lubrication film thickness is particularly sensitive to the number of elements along the vertical direction (ny). Thus, we fix nx=800 in all simulation cases and vary⇥ny⇥to check the grid independence of the solution.…”

Dynamics of droplet breakup in a T-junction

“…In these validation studies, as well as in the breakup study presented here, grid-independent results were obtained with 30 × 30 cells in a square channel cross-section. Details on the numerical methods employed and the validation can be found in Berberovic et al (2009) and Hoang et al (2012).…”

Split of two-phase plug flow with elongated bubbles at a microscale branching T-junction

“…This solver is based on the VOF (volume of fluid) method, and is useful to trace the interface between two immiscible fluids. Therefore, this solver has been adopted to simulate the two-phase flow in microchannels with a reasonable accuracy (Ghaini et al, 2011;Hoang et al, 2012Hoang et al, , 2013Gutiérrez-Montes et al, 2013). The details of the numerical approach are explained in the following section.…”

“…Also, at the channel wall, no-slip and no-penetration conditions were imposed, and a prescribed contact angle in the numerical code was set to zero assuming that a thin liquid film always exists along the periphery of each bubble without having any dry spot (Feng, 2009;Hoang et al, 2012). This assumption was adopted because, according to the experiments of Han and Shikazono (2009) with the similar flow conditions to the present study (the bubble velocity and the capillary number being 0.482 m/s and 0.0063, respectively), the bubble body of about 30 mm in length in a square channel was always surrounded by a thin liquid film.…”