Prediction of Both E-Jet Printing Ejection Cycle Time and Droplet Diameter Based on Random Forest Regression

Area of Science:

• Materials Science and Engineering
• Nanotechnology
• Additive Manufacturing

Background:

• Electrohydrodynamic jet (E-jet) printing is crucial for fabricating flexible electronics and optical devices.
• Accurate control over ejection cycle time and droplet size is essential for E-jet printing quality.
• Establishing precise relationships between printing parameters and outcomes remains challenging due to process complexity.

Purpose of the Study:

• To develop a predictive model for E-jet printing.
• To accurately correlate printing parameters with ejection cycle time and droplet diameter.
• To advance E-jet printing towards precise, drop-on-demand capabilities.

Main Methods:

• Development of a Random Forest Regression (RFR) model.
• Training the RFR model with 72 experimental datasets.
• Utilizing four key printing parameters: voltage, nozzle-to-substrate distance, liquid viscosity, and liquid conductivity.

Main Results:

• The RFR model achieved a Mean Absolute Percent Error (MAPE) of 4.35% for ejection cycle time prediction.
• The RFR model achieved a Root Mean Square Error (RMSE) of 0.04 ms for ejection cycle time.
• The RFR model achieved a MAPE of 2.89% for droplet diameter prediction with an RMSE of 0.96 μm.
• RFR demonstrated superior prediction accuracy compared to CART, SVR, and ANN models with limited data.

Conclusions:

• The developed RFR model effectively predicts ejection cycle time and droplet diameter in E-jet printing.
• This predictive capability offers an efficient method for optimizing E-jet printing processes.
• The model supports the advancement of E-jet printing towards highly accurate, drop-on-demand fabrication.