A two-dimensional CFD model based on volume-of-fluid (VOF) is introduced to examine droplet generation in a cross-junction microfluidic using an open-source software, OpenFOAM together with an interFoam solver. Non-Newtonian power-law droplets in Newtonian liquid is numerically studied and its effect on droplet size and detachment time in three different regimes, i.e., squeezing, dripping and jetting, are investigated. To understand the droplet formation mechanism, the shear-thinning behaviour was enhanced by increasing the polymer concentrations in the dispersed phase. It is observed that by choosing a shear-dependent fluid, droplet size decreases compared to Newtonian fluids while detachment time increases due to higher apparent viscosity. Moreover, the rheological parameters—n and K in the power-law model—impose a considerable effect on the droplet size and detachment time, especially in the dripping and jetting regimes. Those parameters also have the potential to change the formation regime if the capillary number ($Ca$) is high enough. This work extends the understanding of non-Newtonian droplet formation in microfluidics to control the droplet characteristics in applications involving shear-thinning polymeric solutions.

Material extrusion additive manufacturing (ME-AM) techniques have been recently introduced for core–shell polymer manufacturing. Using ME-AM for core–shell manufacturing offers improved mechanical properties and dimensional accuracy over conventional 3D-printed polymer. Operating parameters play an important role in forming the overall quality of the 3D-printed manufactured products. Here we use numerical simulations within the framework of computation fluid dynamics (CFD) to identify the best combination of operating parameters for the 3D printing of a core–shell polymer strand. The objectives of these CFD simulations are to find strands with an ultimate volume fraction of core polymer. At the same time, complete encapsulations are obtained for the core polymer inside the shell one. In this model, the deposition flow is controlled by three dimensionless parameters: (i) the diameter ratio of core material to the nozzle, $d/D$; (ii) the normalised gap between the extruder and the build plate, $t/D$; (iii) the velocity ratio of the moving build plate to the average velocity inside the nozzle, $V/U$. Numerical results of the deposited strands’ cross-sections demonstrate the effects of controlling parameters on the encapsulation of the core material inside the shell and the shape and size of the strand. Overall we find that the best operating parameters are a diameter ratio of $d/D=0.7$, a normalised gap of $t/D=1$, and a velocity ratio of $V/U=1$.

In this paper, the process of encapsulating solid particle(s) into liquid droplets in a high-throughput flow-focusing microchannel is investigated numerically. Open source software is used, which computes fluid flow in an Eulerian framework and particle dynamics with a Lagrangian approach. Previous studies have demonstrated that if no action is taken, particles suspended in a liquid passing through a flow-focusing microchannel will be encapsulated at random. This is perhaps unsurprising, but in one such study, less than 35% of droplets were found to contain exactly one particle. The two aims of this study are (i) to explore the flow patterns arising in a microfluidic channel and (ii) to elucidate the effect of salient governing parameters on encapsulation efficiency (i.e., the fraction of droplets encapsulating one particle) by focusing on ordering the particles before reaching the droplet generation section. Following validation against experimental reference data, the capillary number is varied across the three droplet generation regimes: squeezing, dripping, and jetting. We demonstrate that under certain conditions, an encapsulation frequency of 100% can be achieved with ordered particles, but in most cases, this is significantly lower. We examine the flow field to help understand how this non-uniform distribution of particles occurs. Notably, we find the dripping to be the best option for particle encapsulation and in this case extend the study to explore the effect of junction angle, finding that an angle of 60° is the most favorable. Improved understanding of the encapsulation process derived from this study can help to improve design of high-throughput droplet generation microfluidic systems.

Microfluidics enables generating series of isolated droplets for high-throughput screening. As many biological/chemical solutions are of shear thinning non-Newtonian nature, we studied non-Newtonian droplet generation to improve the reliability of simulation results in real-life assays. We considered non-Newtonian power-law behaviour for Xanthan gum aqueous solution as the dispersed phase, and Newtonian canola oil as the continuous phase. Simulations were performed in OpenFOAM, using the inter foam solver and volume of fluid (VOF) method. A cross-junction geometry with each inlet and outlet channel height (H) and width (W) equal to 50 micrometers with slight contractions in the conjunctions was used to gain a better monodispersity. Following validation of the numerical setup, we conducted a series of tests to provide novel insight into this configuration. With a capillary number, of 0.01, dispersed phase to continuous phase flow-rate ratio of 0.05, and contact angle of 160°, simulations revealed that, by increasing the Xanthan gum concentration (0, 800, 1500, 2500 ppm) or, in other words, decreasing the n-flow behaviour index from 1 to 0.491, 0.389, and 0.302 in power-law model, (a) breakup of the dispersed phase thread occurred at 0.0365, 0.0430, 0.0440, and 0.0450 s; (b) the dimensionless width of the thread at the main channel entrance increased from 0 to 0.066, 0.096, and 0.16; and (c) the dimensionless droplet diameter decreased from 0.76 to 0.72, 0.68, and 0.67, respectively. Our next plan is to study effect of shear-thinning behaviour on droplet generation in different Ca and flow-rate ratios.