Dropwise Condensation Underneath Chemically Textured Surfaces: Simulation and Experiments

Experimental observations of dropwise condensation of water vapor on a chemically textured surface of glass and its detailed computer simulation are presented. Experiments are focused on the pendant mode of dropwise condensation on the underside of horizontal and inclined glass substrates. Chemical texturing of glass is achieved by silanation using octyl-decyl-tri-chloro-silane $(C18H37C13Si)$ in a chemical vapor deposition process. The mathematical model is built in such a way that it captures all the major physical processes taking place during condensation. These include growth due to direct condensation, droplet coalescence, sliding, fall-off, and renucleation of droplets. The effects arising from lyophobicity, namely, the contact angle variation and its hysteresis, inclination of the substrate, and saturation temperature at which the condensation is carried out, have been incorporated. The importance of higher order effects neglected in the simulation is discussed. The results of model simulation are compared with the experimental data. After validation, a parametric study is carried out for cases not covered by the experimental regime, i.e., various fluids, substrate inclination angle, saturation temperature, and contact angle hysteresis. Major conclusions arrived at in the study are the following: The area of droplet coverage decreases with an increase in both static contact angle of the droplet and substrate inclination. As the substrate inclination increases, the time instant of commencement of sliding of the droplet is advanced. The critical angle of inclination required for the inception of droplet sliding varies inversely with the droplet volume. For a given static contact angle, the fall-off time of the droplet from the substrate is a linear function of the saturation temperature. For a given fluid, the drop size distribution is well represented by a power law. Average heat transfer coefficient is satisfactorily predicted by the developed model.