Abstract
Recycling metal powders in the Additive Manufacturing (AM) process is an important consideration in affordability with reference to traditional manufacturing. Metal powder recyclability has been studied before with respect to change in chemical composition of powders, effect on mechanical properties of produced parts, effect on flowability of powders and powder morphology of parts. However, these studies involve ex situ characterization of powders after many use cycles. In this paper, we propose a data-driven method to understand in situ behavior of recycled powder on the build platform. Our method is based on comprehensive analysis of log file data from various sensors used in the process of printing metal parts in the Arcam Electron Beam Melting (EBM) 簧 system. Using rake position data and rake sensor pulse data collected during Arcam builds, we found that Inconel 718 powders exhibit additional powder spreading operations with increased reuse cycles compared to Ti-6Al-4V powders. We substantiate differences found in in situ behavior of Ti-6Al-4V and Inconel 718 powders using known sintering behavior of the two powders. The novelty of this work lies in the new approach to understanding powder behavior especially spreadability using in situ log file data that is regularly collected in Arcam EBM簧 builds rather than physical testing of parts and powders post build. In addition to studying powder recyclability, the proposed methodology has potential to be extended generically to monitor powder behavior in AM processes.
