Acoustic-Based Predictive Quality Monitoring in FDM 3D Printing

This research develops a predictive framework for assessing part quality in Fused Deposition Modeling (FDM) 3D printing using real-time acoustic signals. Conventional quality control methods rely on visual inspection and post-process testing, which are often slow and inconsistent. In this study, low-cost microphones capture the printer’s acoustic emissions during fabrication. Time-frequency features extracted from these signals are analyzed using machine learning algorithms to classify print conditions and detect anomalies such as under-extrusion, layer shifting, or nozzle clogging. Preliminary results indicate that acoustic signatures correlate strongly with process health, enabling early fault detection and automated quality prediction. The approach offers a scalable, non-contact, and cost-effective pathway for real-time quality monitoring in additive manufacturing.