Zinc Oxide Nanoporous Superhydrophilic Surfaces: A Synthesis of Experimental Durability Testing and Droplet Vaporization Model Development Using Machine Learning Methods

Experimental results demonstrate that droplet vaporization on metal surfaces can be significantly enhanced with the application of a nanoporous, superhydrophilic surface coating. A thin layer of ZnO nanopillars can be easily seeded and grown on most metallic surfaces to achieve nanoscale pores between pillars, and ultra-low apparent contact angles. Such surfaces have immense potential to improve spray cooling processes, however, little durability testing of the surface has been performed. In spray cooling applications, as water evaporates, any impurities in the water will be deposited onto the surface. This investigation serves to demonstrate how minerals in hard water deposit on the surface and interact with the ZnO nanopillars of the superhydrophilic surface. Micrographs of the surface demonstrate that minerals deposit nonuniformly, and quickly fill the porous nanostructure. Scale tended to build up on previously deposited scale, leaving largely uncoated areas where droplets chose to preferentially spread, resulting in a continued low contact angle. Maintaining these uncoated areas, and reducing the contaminants present in the water will extend the life and performance of the nanostructured surface. Experiments briefly explored potential chemical cleaning methods but none were found to preserve the nanostructure. It is also important to have a clear model of droplet vaporization on such surfaces and understand the vaporization dependence of surface parameters. Surface and impact parameters such as the surface contact angle, wicking speed and impact velocity all interact to affect the maximum spread of the droplet and the speed at which the droplet reaches its full spread. Along with variations in droplet volume and wall superheat, the model for droplet vaporization becomes more complex and nonlinear. Machine learning tools can be utilized to determine the dependence of droplet evaporation time on these parameters simultaneously. A genetic algorithm and a neural network were used to develop a droplet evaporation model for these superhydrophilic surfaces. Results from the genetic algorithm and neural network were also compared with results from a standard optimization function, the downhill-simplex algorithm. Comparison with the downhill-simplex algorithm is meant to demonstrate the necessity of the other two machine learning techniques in solving this problem. These machine learning techniques were also used to investigate boiling heat flux dependencies of a binary mixture on wall superheat, gravity, Marangoni effects, and pressure. Each algorithm demonstrated clear advantages depending on whether speed, accuracy, or an explicit mathematical model was prioritized. The downhill-simplex algorithm is highly optimized and is most useful when given a highly tailored initial guess. The artificial neural network provides the highest accuracy, but the black box approach prevents the extraction of a simple mathematical model. Finally the genetic algorithm provides sufficiently high accuracy while also being able to produce an explicit model.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.