Browsing by Subject "visual camera images"

Now showing 1 - 1 of 1

In-Situ Quality Inspection of Laser Powder-Bed Fusion Using High-Resolution Visual Camera Images
(University of Texas at Austin, 2016) Aminzadeh, Masoumeh; Kurfess, Thomas
Issues of part quality in terms of quality of fusion and formed porosity are widely known and stated as some of the important challenges with laser powder-bed fusion (LPBF) process. This paper addresses the in-situ inspection of layer-wise part quality using visual camera images. High-resolution visual images are captured from each layer of the part during LPBF process. The imaging and illumination setups are developed such that the produced images visualize detailed surface characteristics of each layer of the build such as fused seams, as well as the individual formed pores. To enable automated inspection of these images, appropriate image processing algorithms are developed to detect individual pores formed in each layer. In addition to detection of individual pores, intelligent pattern matching algorithms are developed, trained, and implemented to identify porous regions from non-porous layers. The surface characteristics of the layers as visualized in camera images can also provide a measure of quality of fusion and the energy of the layer, and an estimated level of porosity. Discussion on characterization of the surface quality in terms of roughness, quality of fusion, and the energy of the build will be made. The results of the automated image analyses provide useful feedback for in-situ process modification as well as part quality assessment.