CORE-SHELL COMPOUND DROPLET IMPACT ON A SOLID SURFACE

Abstract

Droplet impact on solid surfaces controls critical outcomes in printing, coating, spray cooling, encapsulation and additive manufacturing, where both the impact regime and the maximum spreading factor, (i.e., β_max, that is the maximum coverage diameter the drop leaves on the surface normalized by its initial diameter), set footprint, coverage and material retention. In these processes, multicomponent core-shell droplets offer internal structure as an additional degree of freedom, yet their impact behavior is far less understood than that of single-liquid droplets. This thesis investigates the normal impact of millimetric core-shell droplets on smooth solid surfaces, with the aim of predicting impact outcome and β_max using a unified framework that combines experiments, an energy-balance model and machine learning. Experiments are performed with water and water-glycerol cores encapsulated by silicone oil shells of varied viscosity, while systematically varying impact velocity and core volume fraction. High-speed imaging is used to classify impacts into jetting, partial rebound, spreading-contact and their splashing counterparts, and to measure β_max. The data are expressed in a compact multicomponent non-dimensional space based on an equivalent Weber number that accounts for both external and internal interfaces, together with phase-specific Reynolds numbers for core and shell. On this foundation, a compound-droplet energy-balance model is formulated that treats β_max as the outcome of a global balance between initial kinetic and surface energies and viscous dissipation in core and shell, with separate scaling laws for each phase. The model predicts β_max accurately across the experimental parameter range and reveals how energy partitioning shifts with equivalent Weber number, core fraction and shell viscosity exploiting the comprehension of the underlying physics behind the process. In parallel, supervised machine learning models, trained on the same non-dimensional features, deliver high-fidelity predictions of both impact outcome and β_max, while interpretability analyses recover and even improve the dominant physical controls identified by experiments and the energy-balance model. Together, these three pillars demonstrate how internal structure in core-shell droplets can be exploited to manipulate impact behavior and provide predictive tools suitable for analysis and design.