Deep Learning Perspectives on Surface Wettability: LSTM Predictions for Water Droplet Contact Angles

Surface wettability is the ability of a liquid to adhere to a solid surface, determined by the interaction between the liquid and the solid, primarily influenced by the liquid's surface tension and the solid's properties. The surface wettability can distinguish the surfaces into three categories, hydrophilic, hydrophobic, and superhydrophobic. Surface wettability plays various important roles in medical implants, drug delivery systems, and tissue engineering. It also plays an important role in microfluidics, textiles, surface coating materials, and solar panels. The conventional approaches to finding surface wettabilities using experiments may be costly and time-consuming. Hence, this work integrates the pseudo-potential lattice Boltzmann method and deep learning approach for analyzing and predicting surface wettability using solid-fluid interaction parameters (Gads). First, we develop an LBM model for collecting surface wettability data with various parameters. We consider the solid-fluid interaction parameter and droplet radius for analyzing the wettability. We then develop deep learning models using LSTM and do various experiments to find the best models in terms of R
²
score and Mean-Squared Error (MSE) for the performance evaluation.