Surface Quality Characterization and Modeling for Printed 2D Products Using Printing Primitives

Surface quality characterization is essential for guiding design and manufacturing, as well as ensuring product functionality. Although a wealth of literature has been developed for surface quality characterization in mass production, it faces unique challenges in additive manufacturing (AM) due to one-off fabrication of a large variety of products with complex geometries. Surface quality characteristics can vary with built geometries and covariates such as size, built location, and orientation, which makes existing statistical characterization methods not adaptable to frequent design changes in AM. To address these challenges, we proposed a novel surface quality characterization and modeling approach based on printing primitives. We partition a 2D design into a finite amount of printing primitives. The in-plane surface quality patterns on each type of primitives are characterized through tensor-product modeling with integration of domain knowledge to reflect the impact of individual covariates. Simulation studies demonstrate the capability of the proposed model for learning the stochastic process governing the surface quality distribution. We also illustrate the application of the proposed model for analyzing real printed samples produced by an AM process.