Sustained and Enhanced Nucleate Boiling Using Hierarchical Architectures at Large Superheats

Droplet boiling is a common occurrence in many industrial processes, but it can be hindered by the Leidenfrost effect. The Leidenfrost point (LP), defined as the temperature at which an accumulated and stagnant vapor forms between the liquid and the heated solid, consequently deteriorates cooling performance. In this study, inspired by nature, we demonstrate how using a nano-micro hierarchical triple-passage architecture with a higher aspect ratio enhances both vapor and liquid spreading dynamics, boosts heat transfer, and thus elevates the LP. Our results show that the LP is promoted to 273°C, which is a delay of approximately 130°C compared to the LP of 145°C on a copper surface. Through theoretical analysis, we develop a multi-force competition model to reveal the underlying physics of this sustained nucleate boiling. Our findings challenge traditional wisdom, indicating that lower impact velocities of a droplet, though sacrificing the convection, delay the LP through impact pattern manipulation. Additionally, we adopt a physics-informed deep neural network framework to accurately model the nonlinear behavior of droplet boiling (from nucleate boiling to LP) on various surfaces within an ≈11% error. The results here have potential applications in designing more efficient droplet-based boiling heat transfer devices and in controlling droplet boiling at high temperatures.