A New Approach for Online Monitoring of Additive Manufacturing Based on Acoustic Emission

Abstract

Despite its recent popularity, additive manufacturing (AM) still faces many technical challenges for the insufficiency of process reliability, controllability, and product quality. To enhance the process repeatability, effective in-situ monitoring methods for AM processes are needed. In this study, an online monitoring method for AM process failure detection is proposed, where acoustic emission (AE) is applied as the sensing technique. Its application to polymer material extrusion, also known as the technology of fused deposition modeling (FDM), is demonstrated. Experimental results show that the proposed monitoring method allows for the real time identification of major process failures. The occurring time of major failures and failure modes can be identified by analyzing the time- and frequency-domain features of AE hits respectively. A K-means clustering algorithm is applied to verify and demonstrate the classification procedure. The automated failure identification can reduce the waste of fabrication with enhanced machine intelligence.