About this Abstract |
| Meeting |
MS&T25: Materials Science & Technology
|
| Symposium
|
Additive Manufacturing: Equipment, Instrumentation and In-Situ Process Monitoring
|
| Presentation Title |
Mitigating Printing Anomalies in Aerosol Jet Printing: A Data-Driven Approach for Process Planning and Optimization |
| Author(s) |
Shenghan Guo, Hasnaa Ouidadi, Shihab Shakur, Sri Ramesh |
| On-Site Speaker (Planned) |
Shenghan Guo |
| Abstract Scope |
Aerosol jet printing (AJP) is a micro-scale additive manufacturing technology. The use of AJP for manufacturing electronics serving mission-critical needs is hindered by variability in printed features due to prevalent anomalies like line roughness and discontinuity and also overspray. Understanding the origins of these printing anomalies and their relationships with process parameters is an open challenge. This study correlates morphological attributes of printed line and overspray relative to process parameters, establishing a basis for their mitigation. Extraction of the line features and anomalies involves the segmentation of the microscopic images representing the printed lines. Training deep learning (DL) segmentation models requires annotating a large number of microscopic images; a task that is sometimes performed manually and is very time-consuming and tedious. This study uses adversarial domain adaptation to improve the transferability of DL segmentation models between parts printed on different substrate materials,
lowering the need for extensive manual annotation. |