Pattern Recognition-Based Analysis of Free Surface Extensional Flow

Abstract

Understanding and analyzing capillary-driven extensional flow dynamics is essential for applications such as inkjet printing and emulsion formation. However, conventional methods, which often focus on single-point measurements within the slender jet approximation, only partially capture the spatio-temporal evolution of complex fluid shapes during stretching. These shapes, however, contain valuable rheological information. In this study, I introduce novel approaches that integrate machine learning and flow visualization to characterize fluid flows without relying on traditional rheological models. These methods utilize images captured via Dripping Onto Substrate Capillary Break-up Extensional Rheometry (DoS-CaBER), a technique optimized for observing the spatio-temporal dynamics of capillary-driven extensional flows. By analyzing cumulative images and edge curvature, these approaches enable comprehensive extraction of flow information, facilitating an enhanced analysis of the singular dynamics in free-surface extensional flows.