Session: 10.2.3 - Interfacial Phenomena and Flows III

Paper Number: 158132

158132 - Experimental Analysis and Machine Learning Modeling of Maximum Droplet Spreading 

Abstract: 

Droplet impact on a solid surface is a common phenomenon with significant implications for various agricultural and industrial applications. These applications range from hydrophobic surfaces to hydrophilic spray coatings and spray cooling for electronic devices. The spreading behavior of a droplet is influenced by various factors, including its size, impact velocity, density, viscosity, surface tension, contact angle, and surface roughness. However, understanding how these factors interact is complex, making it difficult to accurately predict the maximum spreading. This study examines droplet behavior through experimental visualization and introduces machine learning models to predict the maximum spreading of droplets. The prediction of maximum spreading is based on the Weber number, Reynolds number, and wettability of the liquid. Several well-known machine learning algorithms, including Multiple Linear Regression, Support Vector Machine Regression, Adaptive Boosting Regression, and Extreme Gradient Boosting Regression, were evaluated. Our results show that the Extreme Gradient Boosting regression algorithm achieves an R²-score of 94%, outperforming other models and effectively controls overfitting and underfitting of the data through regularization techniques.

Presenting Author: Jeff Darabi Southern Illinois University Edwardsville

Presenting Author Biography: