Process Monitoring in Metal Additive Manufacturing : Research report of DREAMS project

Abstract

Powder bed fusion process is an additive manufacturing method where material is melted layer-by-layer. Process introduces a lot of heat gradients into the part which influences part quality and tolerances. AM allows manufacturing of complicated geometries, but this makes it difficult to effectively use traditional non-destructive testing methods. Process is also done automatically in an isolated chamber where it is not feasible for the operator to spot caused defects manually. Material is added in AM process which opens new possibilities in the monitoring of the process. Traditional monitoring methods can only monitor what is happening at the surface of the part as when monitoring AM process allows seeing what is happening inside the part when it is forming. Currently there are multiple different methods for process monitoring in PBF. Most common methods are photodiode and visual monitoring due to their cost and ease of use. Optical tomography and thermal imaging are on the rise since there is more information can be gathered from them. Newest method that is still being improved is acoustic since it can accurately detect even small fault if set up correctly. Study was performed on the performance of EOSTATE Meltpool Monitoring (MPM) where models created in such a way that defect would happen. It was noticed that MPM could detect temperature changes in the melt pool and surface quality on xy-plane. Quality on angled could not be detected. Identification of the defect was difficult without addition of visual monitoring and some of the defects could be fixed by earlier layers. MPM is also only focused on the melt pool which means effects on surrounding powder cannot be monitored. Monitoring methods of the PBF can detect potential defect spot with relatively high accuracy. Most methods do require either visual inspection, NDT or destructive testing (DT) to verify the presence of the defect and identification of the defect type.