3D CFD modeling of transient early-stage emulsification in microfluidic cross-junctions: Comparison of URANS and DNS approaches

Water-in-oil (W/O) emulsions are attracting growing interest as alternative fuels, since the water dispersed within the oil promotes secondary atomization through micro-explosions, improving combustion efficiency and lowering pollutant emissions. The quality of such emulsions is largely set during their formation, during the first milliseconds when the two phases meet, yet the initial transient of the process has received comparatively little attention. For this reason, here the early stages of water-in-oil emulsification in a high-throughput asymmetric cross-slot microfluidic device are investigated via transient 3D CFD simulations using complementary URANS and DNS approaches. To the authors' best knowledge, this is the first work to resolve this early-stage, microsecond- scale interface dynamics in such a device, and to cross-validate it across two independent CFD platforms. In the considered device, two inlet streams, water and oil, impinge head-on through square microchannels with cross-sections of 300×300 μm and 600×600 μm, respectively, while the outlet channels have a cross section of 600×600 μm. The models capture the early-stage interface dynamics (first 10–12 ms) in two experimental operating conditions from literature, with respective water flowrates of 7.1 and 15.2 mL/min against 65–67 mL/ min oil, achieving volume fractions of 10–18%. Key results highlight the evolution of local Capillary and Weber numbers on the W/O interface, revealing a transition from incipient threading behavior to stable jetting regime. DNS resolves sharper Ca/We peaks and thinner shear layers than URANS, which smooths gradients. Nevertheless, both approaches result in occupancy ratios in reasonably good agreement with available experimental data and outlet velocities.