Arola, DwayneRamulu, MamidalaSchur, Reid2024-09-172024-09-172024-09-172023Schur_washington_0250E_25323.pdfhttps://hdl.handle.net/1773/52206Thesis (Ph.D.)--University of Washington, 2023Interest in Additive Manufacturing (AM) has quickly progressed from its application to prototypes to stress-critical components. One driving factor is the development of metal AM processes that enable the creation of parts with mechanical properties that rival those achieved by conventional wrought form metal, while enabling nearly unlimited freedom of design. But before widespread adoption of the technology can occur, robust approaches to machine and component qualification must be developed. This “missing link” will require a thorough understanding of the mechanical property variability that exists in the printing of components and its root causes. Variability in the mechanical properties of components produced by metal AM can occur on three hierarchical levels, including intra-build, inter-build, and inter-machine. In this dissertation, each of these sources of variability is considered in Laser Powder Bed Fusion (LPBF) AM of Grade 5 Ti-6Al-4V. To achieve the necessary statistical power, the evaluation is performed through a round robin approach involving six different organizations printing an identical set of six builds on the same make and model printer and with the same lot of gas atomized powder. The intra-build analysis evaluated the spatial variability in microstructure and tensile mechanical properties of metal produced within each of the six printers. While most of the machines exhibited negligible spatial variability in porosity and mechanical properties, the metal from specific machines exhibited parabolic distributions in the porosity and tensile properties. In these machines, the least detrimental porosity (smallest pores overall and fewest number of large pores) was at the center of the build plate. In some machines this trend was oriented along the y-axis (parallel to the gas flow), while in one machine it was radially symmetric in the x-y plane. The shape and spatial position of these distributions suggest that gas flow and laser incidence angle are the most likely root causes to variability. The inter-build variability was evaluated in terms of the porosity and tensile properties with respect to powder reuse and changes in machine health. A careful analysis of the powder chemistry and particle size distribution showed that the powder quality did not degrade significantly with reuse over the six builds. Furthermore, there were no trends in porosity or tensile properties due to powder reuse. However, the strength of metal produced from one machine in the round robin did undergo degradation related to machine health. Although a definitive root cause was not identified, a change to the laser spot size occurred, which could have caused the decrease in tensile properties. Finally, the inter-machine variability was also evaluated through the metal porosity and mechanical properties. There were significant differences in porosity and tensile properties between many of the machines. While the porosity distributions were statistically similar for five of the six machines, lack of fusion (LOF) pores were found in metal of the sixth machine, with as many as 0.3 pores/mm3 and diameter ≥ 0.125 mm. The largest variability in tensile properties resulted from differences in the post-processing heat treatments. For the machines with identically heat treated metal, the properties exhibited coefficients of variation (CoV) in strength that rivaled that of wrought form titanium (1-2%). The largest CoV was exhibited in the strain at failure, with values as large as 12.6%. Results from this round robin investigation are the first to quantify the variability in mechanical properties within and across multiple identical machines operated in different organizations. As such, the results and complimentary analyses make a unique contribution to the literature and establish a foundation of understanding that supports industrialization of LPBF. This work will support efforts aimed at machine qualification and the certification of components. However, further research on the root causes of process variability and its effect on fatigue properties will be a critical next step to the widespread adoption of the technology for aerospace applications.application/pdfen-USnoneAdditive manufacturingLaser powder bed fusionMechanical performancePorosityQualificationTitaniumMaterials ScienceMechanical engineeringMaterials science and engineeringThesis