In Situ, Nondestructive Inspection During Additive Manufacturing of Metallic Parts

The research deliverable of this project will be a prototype hardware-software system for layer-by-layer nondestructive inspection for near-real time determination of spatial distribution of the probability of material defects as a metallic part is being fabricated using the selective laser melting (SLM) process. The hardware will comprise an off-the-shelf laser scanner (profilometer) and precision linear motion, which will be implemented into the deposition chamber of commercial SLM equipment. The software will comprise signal processing algorithms for data reduction and the processing of the collected 3-D top surface build geometry / roughness data for the determination of the probability of material defects. The hardware-software prototype will collect and process the data in a time scale that causes minimal impact on the operational efficiency of the SLM process. There are five technical objectives: (1) Collect layer-by-layer build geometry and powder recoat data; (2) Determine the spatial distribution of the probability of material defects; (3) Characterize the probability of detection (POD); (4) Utilize physics-based models to correlate the phenomenology of interaction between surface roughness and material defects; and (5) Demonstrate prototype of the voxel-by-voxel Anomaly Flagging System.