ConvNet-based prediction of droplet collision dynamics in microchannels

The dynamics of droplet collisions in microchannels are inherently complex, governed by multiple interdependent physical and geometric factors. Understanding and predicting the outcomes of these collisions (whether coalescence, reverse-back, or pass-over) pose significant challenges, particularly due to the deformability of droplets and the influence of key parameters such as the density and viscosity of immiscible fluids, the initial offset between droplets, and the confined geometry of microchannels. Traditional methods for analyzing these collisions, including computational and experimental techniques, are time-consuming and resource-intensive, limiting their scalability for real-time applications. In this work, we explore a novel data-driven approach to predict droplet collision outcomes using convolutional neural network (CNN). CNN-based approaches present a significant advantage over traditional methods, offering faster, scalable solutions for analyzing large datasets with varying physical parameters. Using a lattice Boltzmann method for binary immiscible fluids, we numerically generated droplet collision data under confined shear flow. These data, represented as droplet shapes, serve as input to the CNN model, which automatically learns hierarchical features from the images, allowing for accurate and efficient collision outcome predictions based on deformation and orientation. The model achieves a prediction accuracy of 0.972, even on test datasets with density and viscosity ratios not included in the training. Our findings suggest that the CNN-based models offer improved accuracy in predicting collision outcomes while drastically reducing computational and time constraints. This work highlights the potential of machine learning to advance droplet dynamics studies, providing a valuable tool for researchers in fluid dynamics and soft matter.