Event-driven Simulation of Multi-scale Dropwise Condensation

Abstract

Simulation of the multi-scale dropwise condensation on hydrophobic/superhydrophobic surfaces is of great significance for exploring condensation mechanism and guiding related engineering applications. However, the computational time cost of the conventional time-driven simulation is very expensive because the dropwise condensation covers multiple temporal and spatial scales and involves various behaviors between droplets. Here, an event-driven simulation method is proposed to efficiently simulate the multi-scale three-dimensional dropwise condensation with its simulation accuracy verified against the theoretical model and experimental results in literature. The dynamic octree data structure and its related algorithms are adopted to improve computational efficiency. Compared with the two-dimensional time-driven simulation, the computing efficiency of the three-dimensional event-driven simulation is improved by one to two orders of magnitude. The influence of droplet departure modes on dropwise condensation is then discussed and the results show that when droplets mainly depart through droplet jumping instead of droplet sliding, the surface coverage rate is smaller but the droplet population density and heat flux are larger. The increase of subcooling degree and nucleation density, or the decrease of critical sliding and jumping radius, can promote droplet departure and enhance condensation heat transfer, and the contribution of droplet jumping to heat flux is dominant. This work provides a new method that is efficient and reliable for multi-scale simulation of dropwise condensation and advances the understanding of dropwise condensation.