Near real-time geometric error detection via in situ layer-wise 3D scanning and signed distance evaluation

Additive manufacturing’s (AM) layer-wise construction is susceptible to process variation and fabrication errors. In-situ monitoring techniques have been proposed to detect errors, but they limit inspection to a localized deposition, a single layer, or to a final part state, and fail to capture how layers change throughout processing. Embedding 3D scanners into the AM build volume could address these limitations. However, conducting layer-wise evaluation of high-resolution scan data presents sensing, geometric analysis, and data management challenges. To address these limitations, a novel methodology is presented to rapidly capture, characterize, temporally merge, and quantify geometric error of a printing part using embedded 3D scanning, a temporal merging algorithm, and a coupled efficient computational geometric analysis (adaptively sampled distance function). It is found that in-situ geometric deviation of printed structures can be analyzed within seconds over relatively large build volumes (e.g., 350 cm 3 ) at a high spatial resolution (e.g., 0.3 mm). The proposed method is agnostic to any AM process in which the full part is visible throughout the build (e.g., material extrusion, directed energy deposition). In this work, the method is validated via a layer-by-layer analysis of directed energy deposition arc (DED-Arc) AM.