Defects, Process Parameters and Signatures for Online Monitoring and Control in Powder-Based Additive Manufacturing

Abstract

Additive Manufacturing (AM) is a process that is based on manufacturing parts layer by layer in order to avoid any geometric limitation in terms of creating the desired design. In the early stages of AM development, the goal was just creating some prototypes to decrease the time of manufacturing assessment. But with metal-based AM, it is now possible to produce end-use parts. In powder-based AM, a designed part can be produced directly from the STL file (Standard Tessellation Language/ stereolithography) layer by layer by exerting a laser beam on a layer of powder with thickness between 20 μm and 100 μm to create a section of the part. The Achilles’ heel of this process is generation of some defects, which weaken the mechanical properties and in some cases, these defects may even lead to part failure under manufacturing. This prevents metal-based AM technology from spreading widely while limiting the repeatability and precision of the process. Online monitoring (OM) and intelligent control, which has been investigated prevalently in contemporary research, presents a feasible solution to the aformentioned issues, insofar as it monitors the generated defects during the process and eliminates them in real-time. In this regard, this paper reveals the most frequent and traceable defects which significantly affect quality matrices of the produced part in powder-based AM, predominately focusing on the Selective Laser Sintering (SLS) process. These defects are classified into “Geometry and Dimensions,” “Surface Quality (Finishing),” “Microstructure” and the defects leading to “Weak Mechanical Properties.” In addition, we introduce and classify the most important parameters, which can be monitored and controlled to avoid those defects. Furthermore, these parameters may be employed in some error handling strategies to remove the generated defects. We also introduce some signatures that can be monitored for adjusting the parameters into their optimum value instead of monitoring the defects directly.