Abstract
The lack of direct process control of silicone material extrusion (MEX) limits the accuracy and design complexity of printed structures. In this paper, an iterative process monitoring system for silicone MEX that enables in-situ quality monitoring and optimization is presented and validated. To achieve in-situ quality monitoring, a laser profile sensor is integrated into the 3D printing platform to measure dimensional errors. The process includes virgin printing, scanning, analyzing, judging, and modified printing. Strands are initially printed and scanned step by step with the coordination of motion and sensing systems to obtain a point cloud with 3-dimensional coordinates. Data are corrected, standardized, and analyzed section by section instead of using an image process to improve efficiency and reserve information in the height direction. Centre, width, and height information are extracted for each strand cross-section. Noise and anomalies are then removed by a clustering algorithm for the center points to obtain the actual trajectory centerline and distribution of strand width along the trajectory. To achieve in-situ optimization for errors in trajectory, strand width, and height, the trajectory and process parameters are modified in the next print. The trajectory is corrected by applying the opposite trajectory deviation error vector. The feed rate is modified according to our previously proposed strand profile model that maps the process parameters to the strand cross-section geometry parameters. Compared with direct printing without in-situ quality monitoring and optimization, the overlap between the actual and the target strand was improved by 18%. The presented method demonstrates the feasibility and capability to improve accuracy and reduce errors.
