Surface Wettability Modeling and Predicting via Artificial Neural Networks

Area of Science:

• Materials Science
• Surface Science
• Computational Modeling

Background:

• Surface wettability, quantified by contact angle, is critical for liquid-solid interactions in fields like lubrication and heat transfer.
• Liquid behavior on surfaces is governed by surface texture and chemical properties.
• Accurate prediction of surface wetting is essential for optimizing material performance.

Purpose of the Study:

• To model and predict surface wettability by adjusting surface texturing parameters.
• To develop and optimize artificial neural network (ANN) models for contact angle prediction.
• To identify the most effective ANN configurations for accurate wettability forecasting.

Main Methods:

• Utilized artificial neural networks (ANNs) to predict surface contact angles.
• Systematically varied ANN parameters, including activation functions, hidden layer count, and training algorithms.
• Evaluated model performance using the coefficient of determination (R²).

Main Results:

• Achieved a high prediction accuracy for contact angles, with R² values exceeding 0.9.
• Demonstrated the capability of ANNs to model surface wettability based on surface parameters.
• Identified optimal ANN configurations for predicting surface wetting behavior.

Conclusions:

• Surface texturing parameters can be effectively adjusted to model and control surface wettability.
• Artificial neural networks provide a powerful tool for predicting contact angles and surface wetting.
• The developed ANN model offers a reliable method for forecasting wettability in diverse applications.