Surface quality prediction for FDM-printed parts using in-process material flow data

Abstract

Efficient and sustainable manufacturing practices rely on the early detection and removal of faulty components in production processes. In the context of Fused Deposition Modeling, this means that identifying defective parts during the production cycle can help to minimize waste and optimize resource utilization. However, conventional quality control methods, which involve post-process inspection, can be time-consuming and inefficient, particularly if nonconforming parts are detected after the production cycle is complete. To address this issue, a real-time quality prediction system has been developed that utilizes in-process flow sensor data to detect and identify nonconforming parts as they are being produced. The system was tested on cuboid test specimens, which were deliberately modified to include defects on the surface of the part. By analyzing the sensor data in real-time, the system was able to identify the defective parts and provide corrective actions to minimize waste and optimize resource utilization. By implementing this approach, manufacturing processes can be streamlined and resource utilization can be optimized while minimizing the production cycle time. This approach represents a significant advance over traditional quality control methods, which rely on post-process inspection and human factors.