Intelligent Optimization Method of Piezoelectric Ejection System Design Based on Finite Element Simulation and Neural...

Area of Science:

• Additive Manufacturing
• Materials Science
• Computational Fluid Dynamics

Background:

• Piezoelectric ejection systems are crucial for high-resolution additive manufacturing.
• Optimizing these systems is complex, requiring reliable prediction of droplet ejection behavior.
• Conventional design methods can be time-consuming and less accurate.

Purpose of the Study:

• To develop an intelligent method for modeling and optimizing piezoelectric ejection system design.
• To enhance the speed and reliability of designing piezoelectric ejection system parameters.
• To accurately predict droplet ejection behavior (DEB) for improved additive manufacturing processes.

Main Methods:

• Developed a finite element (FE) model of the droplet ejection process, validated with experimental and literature data.
• Created and optimized two artificial neural network (ANN) models using feed-forward back propagation and genetic algorithm (GA).
• Established a database via FE calculations to train ANN models correlating design parameters with DEB indicators (jetting velocity, droplet diameter).

Main Results:

• ANN models accurately predicted droplet jetting velocity and diameter using training and testing data.
• The combined FE simulation and ANN-GA approach determined optimal piezoelectric ejection system designs.
• Experimental validation showed low prediction errors (4.48% for jetting velocity, 3.18% for droplet diameter).

Conclusions:

• The intelligent method provides a reliable and robust approach for piezoelectric ejection system design optimization.
• This integrated simulation and AI technique significantly improves the accuracy and efficiency of additive manufacturing system design.
• The study validates the predictive power of the developed models for critical droplet ejection characteristics.