Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Process data-based knowledge discovery in additive manufacturing of ceramic materials by multi-material jetting (CerAM MMJ)

: Lang, Valentin; Weingarten, Steven; Wiemer, Hajo; Scheithauer, Uwe; Glausch, Felix; Johne, Robert; Michaelis, Alexander; Ihlenfeldt, Steffen

Fulltext ()

Journal of manufacturing and materials processing 4 (2020), No.3, Art. 74, 16 pp.
ISSN: 2504-4494
Journal Article, Electronic Publication
Fraunhofer IKTS ()
Fraunhofer Singapore ()
data management; design of experiments; multi-material jetting; ceramics; additive manufacturing; Lead Topic: Digitized Work; Research Line: Machine Learning (ML); 3D printing; knowledge acquisition; materials research

Multi-material jetting (CerAM MMJ, previously T3DP) enables the additive manufacturing of ceramics, metals, glass and hardmetals, demonstrating comparatively high solid contents of the processed materials. The material is applied drop by drop onto a substrate. The droplets can be adapted to the component to be produced by a large degree of freedom in parameterization. Thus, large volumes can be processed quickly and fine structures can be displayed in detail, based on the droplet size. Data-driven methods are applied to build process knowledge and to contribute to the optimization of CerAM MMJ manufacturing processes. As a basis for the computational exploitation of mass sensor data from the technological process chain for manufacturing a component with CerAM MMJ, a data management plan was developed with the help of an engineering workflow. Focusing on the process step of green part production, droplet structures as intermediate products of 3D generation were described by means of droplet height, droplet circularity, the number of ambient satellite particles, as well as the associated standard deviations. First of all, the weighting of the factors influencing the droplet geometry was determined by means of single factor preliminary tests, in order to be able to reduce the number of factors to be considered in the detailed test series. The identification of key influences (falling time, needle lift, rising time, air supply pressure) permitted an optimization of the droplet geometry according to the introduced target characteristics by means of a design of experiments.