Image-based analysis of component geometry throughout the binder jetting printing process

A high scrap rate caused by quality shortcomings of the components hampers the binderjetting process in terms of its economic viability. In an effort to address this issue and ensureconsistent and high quality parts, this work introduces binder placement accuracy monitoringin the printing process. For this purpose, the Contact Image Sensor(CIS) unit of a commercialscanner was used to generate images of the powder bed with 2400 dpi, which depict the binderplacement in high resolution. An algorithm was proposed to evaluate those images, with theresult that different edges of the binder insertion were successfully detected. A constant errorin binder insertion was observed over a binder saturation range of 20-125 %. This observationis contrasted by the geometric evaluation of the cured components, whose volumes werealso determined and computationally illustrated. Here, a saturation-like progression of thevolumes could be observed with increasing binder saturations. The volume increase at lowbinder saturations can be explained by the insuffcient bonding of the powder particles, whichdetach from the green part during the depowdering and thus explain the volume loss. Anincrease in volumes from the saturation curve is observed at high binder saturations and is dueto the occurrence of bottom bleeding. As benchmarks, ImageJ and caliper measurements wereperformed, confirmin the algorithm results. The establishment of further error detectionsand the integration of the system into the ongoing printing process are necessary before animplementation of process control for quality assurance can be successfully.