Mass finishing for additive manufacturing: Tribological analysis, surface topology, image processing, predictive model, and processing recommendations

Although additive manufacturing has been explored for diverse applications, poor surface finishes are common to as-built AM parts. Particulary in metal AM, often post-processing is required to improve surface quality. Centrifugal disc finishing (CDF) is an efficient mass finishing technique that is widely used in traditional manufacturing methods. As-built AM surfaces have unique surface topology based on build orientation that respond differently to mass finishing conditions. This work investigates the effects of AM processing and CDF parameters on evolution of surface topology based on process conditions . Prediction accuracy of 90% was achieved in surface evolution using feature-instance segmentation and machine learning method implemented in this study. The study provides a new insight into achieving desired surface finish in metal AM for complex geometries via mass finishing such as centrifugal disk finishing. Surface topology and contact mechanism-based AM+CDF manufacturing recommendation for different AM build orientations. • A novel image processing method for AM surfaces is to quantify build orientation effects is presented. • Key contact mechanisms of material removal and material distribution have been identified based on AM build orientation. • Achieved 90 % accuracy in prediction of surface roughness evolution as a function of build orientation and processing conditions (time and speed) using a machine learning model. • Established processing guidelines for surface finishing conditions tailored to AM surface orientations.