Multiscale Investigations of Powder Morphology on EBM Product Quality

Abstract

Electron beam melting manufactures parts using a fine metallic powder making the process applicable to thin structural features. While it is generally accepted that as-built part quality is a function of feedstock quality, little has been done to understand how particle morphology contributes to these relationships. Therefore, the objective of this work was to investigate particle-process-part quality relationships in electron beam melting for reused Ti-6Al-4V powder. This was accomplished by developing experimental and numerical methods to study the four process stages: 1) Particle characterization; 2) Feedstock behavior; 3) Thermal interactions; and 4) As-built part quality. In general, particles were classified as: 1) Circular; 2) Fractured, Dented, Deformed, or Elongated (FDDE); or 3) Satellites, Sintered, Agglomerated, or Molten (SSAM) using optical microscopy. Feedstock behavior (e.g., flowability; packing density; angle of repose) was evaluated under atmospheric “benchtop” and vacuum “machine” conditions using a custom electromechanical Hall flowmeter stand. A “Start Plate Probe” was also developed to understand how start plate setup contributed to increases in local layer thickness and packing density. Additionally, powder spreading behavior was examined numerically using the discrete element method, the results of which were correlated with experimental measurements. Finally, thermal interactions and as-built part quality (e.g., microstructure/phases; defects) were studied using several numerical (e.g., volume of fluid; element birth-death; phase field) and experimental (e.g., optical microscopy; scanning electron microscopy; x-ray computed tomography) techniques. The results indicate that relative percentages of the circular, FDDE, and SSAM particles were found to be 77:10:13%. The high circularities of FDDE particles and low percentages of SSAM particles were found to not affect flow and packing behavior in a statistically significant way. Likewise, a comparison of testing within atmospheric and vacuum environments indicate that differences in flow and packing behavior under these conditions is not statistically significant. Alternatively, benchtop measurements are a good indicator of powder performance within vacuum environments at room temperatures. However, the results said nothing about the high-temperature behavior which is expected to be significant due to sintered neck formation and fracture. During powder raking, quality of the powder bed as measured by packing density was found to be strongly correlated with rake-start plate clearance as a result of particle jamming. Results for thermal interactions were well correlated with experimental measurements for ⍺ lath lengths, prior-β widths, and beam penetration estimates. Finally, good correlations were found between the relative sizes of SSAM particles and lack of fusion defects in as-built components indicating a potential mechanism for defect formation.