This thesis presents a methodology for fixturing, scanning, and orienting additively manufactured (AM) metal parts prior to finishing operations using a process called CNC-RP (rapid prototyping), which is a subtractive rapid machining method that employs the concept of sacrificial supports. AM has enabled the manufacturing of complicated designs in a layer-based fashion but often produces parts with inadequate surface roughness and/or dimensional control. Subtractive manufacturing (SM) techniques typically fabricate parts with overall better surface roughness and feature accuracy but require fixtures and custom tooling for each part design. Combining the two technologies of additive and subtractive manufacturing to create metal parts would allow the design flexibility of additive while simultaneously enabling the accuracy and finish of subtractive. However, the process of locating components when two processes are involved can be challenging. Solutions to this challenge have been applied to the metal casting industry in the form of custom fixtures, but the fixtures made for these finishing operations are intended for high volumes of a single part; thus, these fixtures are too expensive to justify for the small batch sizes made through additive manufacturing.

Through non-contact scanning, a set of algorithms built into a localization software program identifies the location and orientation of a part secured within the fixture and outputs this identification in the form of a position vector. These algorithms also yield the angular rotations required to align the part's current position vector to the ideal computer aided design (CAD) model position vector in preparation for CNC-RP. Another program was developed to determine the finite rotations of the A- and B-axis when part length is taken into consideration. These rotations were physically implemented with a fixture hardware element.

Process capability metrics were employed to validate this method. For both axes, parts could be produced within tolerance only after this method was employed (Cp and Cpk values greater than 1.0 are “acceptable”); however A-axis correction fell short with a Cpk value of 0.985, but tighter process control and/or more accurate equipment may resolve this issue. Thus, this thesis provides a feasible methodology to combine additive manufacturing and CNC-RP subtractive manufacturing.