Computer Tomography based improvement of in-process monitoring capabilities

Autonomously correlate in-line CT data with AM process parameters decided on during build preparation. To quantify variations in the process, all experiments will be monitored in-line using, e.g., thermal and optical imaging. According to an appropriate data model, the monitoring data will be stored in a database, and the manufactured parts will be CT scanned. The monitoring data will also be acquired for every process layer and stacked to produce a 3D volumetric model. Deep learning techniques will then correlate the in-process monitoring models with the post-process CT data. The scope of this correlation is to identify the sensor output data range that yields adequate part quality. Combining the CT results with the monitoring data, the CT scans of the experimental parts will be interpreted to develop predictive models to be used in build preparation.