Date of Award


Degree Name

Master of Science




Ryan Wicker


At the writing of this Thesis, additive manufacturing (AM) also known as 3D printing, has been popularized for its diversity in applications ranging from home and personal use, medical, industrial, consumer products, aerospace, architecture, automotive, military, fashion, food, art and more. Industries taking advantage of the design freedom and complexity offered by AM have exploded the growth of the technologies. Specifically, technologies that process metals using electron and laser beams have been recognized by the aerospace industry as a promising avenue for re-engineered components leading to reduced weight and improved engineering efficiencies for components like engine brackets and nozzles. However, the direct implementation of AM has not been straightforward primarily because AM processes are not fully trusted to produce reliable and reproducible parts. Continued research, including the subject of this Thesis, is aimed at better understanding process variations in the AM of metals via powder bed fusion.

Electron Beam Melting (EBM) is an AM is a technology in the category of powder bed fusion that is seeing increased adoption by a variety of industries for part production. EBM was the focus of this study for the fabrication of solid and porous parts using precursor powder composed of Ti-6Al-4V. This research is centered on evaluating mechanical properties, analyzing microstructures, and correlating the fabrication process to inherent characteristics of solid and porous parts fabricated utilizing EBM. For dense parts, data were documented on the effects on mechanical and microstructural properties from neighboring parts and building location. In the case of porous, or lattice structures, data were documented on the effect of parameter modifications, such as processing currents and number of scan passes, on the final part mechanical response and microstructure. Solid and lattice components were mechanically tested and microstructural features were obtained by the use of computer software MATLAB. Microstructural and factographic analyses were performed on samples prepared for such analysis and the prepared surfaces were used for hardness testing.

Several variables in part production exist for AM, including, but not limited to, part orientation, part location, processing parameters, and geometry. For the solid components fabricated for this Thesis, the core objective was to determine the effect that surrounding parts have on mechanical properties. The study was expanded to also determine how part location within a build area affects the mechanical properties. Mechanical information of a material was obtained by tensile testing that provided values for the following properties: yield strength (YS), ultimate tensile strength (UTS), modulus of elasticity (E), and elongation (%EL). Twenty-seven total parts were fabricated, machined, and tested. Further comparison between build characteristics and microstructure was evaluated by performing metallographic analysis. The surfaces prepared for metallographic analysis were subsequently subjected to hardness testing and fractography analysis was performed as a means to construct a correlation between the gathered information and the potential cause(s) contributing to the ductile failure mode observed.

Results from surface area variations, in general, demonstrated change in YS, UTS, and E. Percent elongation reflected significant improvements from the increased surface areas. Noteworthy was the significant decrease in standard deviation as melt surface area increased. The typical Widmänstatten lath morphology was observed by all EBM-fabricated Ti-6Al-4V parts; however, increases in α lath width were evident from the increased surface area. For the study consisting of spatial distribution within a build, in general, an increase in YS, UTS and E was observed for parts located toward the back of the fabrication system, with respect to the front of the system. Percent elongation decreased from the back to the front of the machine.

The fabrication of generic structures provided the information on features produced using the standard commercial methodology, which would generally be used by industry. One of the benefits provided to Arcam users is the ability to access the process menu for customize build parameters. The previous deigns presented a study on features for solid structures fabricated without processing parameter variations. This furthering task presents for modification of processing parameters on lattice structures. The area that has been studied adds to improvements on features by additive manufacturing in general, enabling sophisticated designs with tailored properties while optimizing material and weight.

A particular focus of this research was the study of lattice structures fabricated by EBM. Mechanical information of a material was obtained through compression testing, which provided values for the following properties: youngâ??s modulus of elasticity (E), ultimate compression strength (UCS), the fracture load, and the displacement seen at such. Twenty-seven total lattice specimens were fabricated and tested. A correlation between microstructure and properties was explored using metallography analysis. Samples that displayed considerable property differences within a single build were chosen for analysis and consequently subjected to hardness testing. Fractography analysis was also performed on selected specimens to examine the potential cause(s) contributing to the bimodal failure mode.

Of pertinence to lattice structures is removing the detrimental martensitic phase created during rapid solidification, which causes brittle mode fracture. Lattice structures provide a light weighting alternative to the typically utilized solid implant that will offer physical properties compatibility. Implants are just one example of the products lattice structures can be utilized for, however the possibilities include the industries already mentioned . The structures cannot be included in production unless performance away from catastrophic failure, as the one the martensitic phase mentioned would promote is ensure. In order to change the microstructure of lattice structures, it was necessary to make parameter modifications on the fabrication process. The processing currents studied for comparison included a standard build and three sets of modified scan speeds and currents. A second investigation involved the mechanical response of increasing layer-by-layer scans. The melts studied included a single scan, a double scan, and a triple scan. In the case of the study consisting of varying processing speeds and currents, in general, the results showed a change in properties from the standard because of parameter modifications. The E and UCS values do not change notably and fracture load observed varied considerably. A decrease in α lath width was observed as processing current decreased. In the case of the study consisting of increasing melts, in general, the change in mechanical properties showed improvements across all parameters of E, UCS, and fracture load. Microstructurally, this research demonstrated the removal of the detrimental martensitic phase in lattice structures by altering the thermal environment within the build chamber (through changing scanning strategies) during the build. The study conducted in this Thesis achieved the reengineering of the microstructure observed in lattice components to a complete Widmänstatten lath morphology with the appearance of α equiaxed grains.

Overall, the research outcome of this Thesis provided further characterization of the EBM fabrication process and presented some potential improvements to unique lattice structures. The mechanical response obtained by increase of surface area and varying locations showed that certain variations are present and users need to be aware of these part-to-part differences. In the case of components containing lattice structures, the mechanical response obtained by variation of the processing current within a single build and increasing scan passes suggested that detrimental martensite phase that typically occurs in the standard build process can be removed from in-process modifications.




Received from ProQuest

File Size

129 pages

File Format


Rights Holder

Paola Azani