Infrared-assisted acoustic emission process monitoring for additive manufacturing

Abstract

A known limitation of additive manufacturing (AM) with laser powder bed fusion machines is the long build time. Real-time monitoring of the additive process can provide an early alert if the process exceeds preset limits of the processing parameters. This has a significant advantage over conventional post-manufacturing nondestructive testing methods. In-process monitoring can be used with feedback to either start a set of corrective actions or terminate the build early to save machine time and powder. One of the monitoring technologies that has been investigated by several research groups is acoustic emission (AE). The signals from the AE sensors can be correlated with several parameters of the AM process such as laser power and volumetric energy deposition. The difficulties exist in interpreting the AE signals. For example, the layer-to-layer geometry variations or hatching pattern variations can produce relatively large changes in magnitude or frequency of the AE signals that could be misinterpreted as a process abnormality. This work describes an investigation conducted at CCAM for the purpose of improving the interpretation of the AE signals. To do this, infrared thermographic imaging was used to better interpret the cause of the AE signal variations. A low cost infrared camera with a micro-bolometer sensor was used to acquire a series of infrared images during the AM process, in addition to recording the AE signals. A comparison between the position of the laser beam on the part and the AE signals resulted in understanding conditions under which the AE signal was produced. This provided additional input to the event interpretation process. The combined application of these two technologies resulted in a better separation of the AE signals into clusters related to a particular change in the AM process recipe. Future work will be devoted to combining infrared and AE data to reduce signal deviations caused by the change in the laser beam pattern. This will improve reliability in detection process abnormalities that reduce quality of the additive components.