Abstract Scope |
Aerosol jet printing centers on the deposition of atomized materials. The process can fabricate high-resolution features, for the creation of microelectronics, flexible electronics, and biomimetic surfaces, among others. However, deposit quality is a complex function of the process parameters (sheath gas flow, carrier gas flow, and print speed). Improper selection of process parameters or process drifts often leads to suboptimal print quality and the potential to alter the print’s functionality. Accordingly, the objective of this work is to ensure flaw-free deposition. To realize this objective, we use machine learning strategies to classify and predict the characteristics of the deposit as a function of the process parameters and theoretical model results. Once trained for a machine setup and material, the resulting machine learning network can identify the deposit quality as a function of the process parameter combination, prior to printing, saving operator time on failed prints. |