Modeling of the shape of Taylor cone in EHD printing based on backpropagation neural network and genetic algorithm

Abstract

Electrohydrodynamic (EHD) printing has been recognized as a promising additive manufacturing technology with superior pattern resolution and economic viability. The shape of Taylor cone is deemed a pivotal element for the amplification of deposition efficiency, and the maintenance of consistent operational steadiness in EHD printing. The correlations between diverse operating parameters and the shape of Taylor cone are presently not well investigated. In this paper, modeling of relationships between operating parameters and the shape of Taylor cone was conducted with a backpropagation neural network (BPNN) and a genetic algorithm optimized backpropagation (GA-BP) neural network. Taylor cone semi-vertical angle and the meniscus height were employed as two indexes to characterize the shape of Taylor cone. The prediction accuracies of BPNN model and GA-BP model were 92.79 % and 95.46 %, respectively. The GA-BP model demonstrated higher precision in forecasting the shape of Taylor cone. A predictive framework for the shape of Taylor cone was proposed, which provided a practical tool for process optimization in EHD printing.