In additive manufacturing, laser powder bed fusion (LPBF) has unrivaled strengths due to its design and manufacturing freedom. The in situ validation of additively manufactured components would reduce or entirely remove the need for post-processed non-destructive evaluation. Potentially enabling the direct utilization of components from the print bed. However, typical approaches to in situ monitoring of the LPBF process utilize high-speed thermal and optical cameras coupled with advanced optics to enable co-axial imaging of the weld pool. The amount and quality of the data obtained through these systems necessitate the need for extensive post-processing of data. In contrast, this work provides a low-cost in situ monitoring and real-time computing alternative using industrial cameras and optical filters to track the splatter area of the welding process. To reduce the dimensionality of data retained for a given component, the proposed process tracks the brightness contours of the welding process in real-time and retains only a select number of features. In this introductory work, the prototype system is investigated using a variety of different image processing methods to optimize processing speed (measured in frames per second) versus the size of melting splatter for a test specimen of 10 mm × 10 mm × 5 mm. Defects in the specimen are quantified using computed tomography and linked to information extracted from tracking the splatter-related features in situ. Results show that the speed of the computational system, visibility of splatter, and the accurate translation of splatter brightness to contours with area and locations is critical to functionality. A discussion on the trade-offs between these constraints is provided.
Improvements to processes and materials have led to increased additive manufacturing capabilities using the fused filament fabrication method in terms of speed, quality, and repeatability. However, there are significant challenges in guaranteeing the desired output quality due to uncertainties inherent to the printing process. These include uncertainties in the quality of raw materials across different batches, fabrication environment (e.g., humidity, temperature), and machine wearing. The widespread adoption of fused filament fabrication for industrial applications faces considerable challenges in reducing part-to-part variations and assuring the mechanical properties of a manufactured component. In this paper, an in situ fault detection platform that considers the structural properties of the printed part is proposed. The presented system uses the optical camera and a deep learning methodology to detect faults online using training sets developed offline. The performance of the system is quantified using a variety of metrics. Computational speed for inference computation, minimum fault-sized detection, and measurement noise in the system are examined in this work.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.