Dense emulsion flows in confined geometries are pivotal across diverse natural, biological, and industrial processes, yet modelling such flows faces the key challenge of accurately simulating non-coalescent behaviour of near-contact dispersions. In this work, we develop a novel repulsive force model within the colour-gradient lattice Boltzmann framework, incorporating dispersion densities and separation distance to model near-contact interactions, thus enabling the simulation of non coalescent behaviour. After validating this model by simulating binary droplet bouncing and soft crystal formation, we use it to investigate emulsion rheology in 3D Poiseuille flow, analysing effective viscosity μe dependencies on dispersion volume fraction Φ and capillary number Ca across various viscosity ratios m. Results show an exponential μe increase with Φ under moderate Ca and m, matching theoretical predictions. Velocity profile and viscous dissipation analyses reveal that dispersion induced viscous dissipation governs μe enhancement, though strategic repositioning dispersions away from walls reduces μe. By varying dispersion viscosity, we find that for m ≥ 1, μe -Φ curves follow established trends, while for m ≤1:20, the counterintuitive μe < 1 phenomenon emerges, representing the first demonstration of viscosity reduction via low-viscosity dispersions. Finally, the emulsion system exhibits shear-thinning behaviour at high Φ, as increasing Ca drives centerward migration and elongated deformation of dispersions, thereby reducing hydrodynamic resistance. These results provide new fundamental insights into controlling emulsion rheology through dispersion engineering.

无