Predicting Process and Material Design Impact on and Irreversible Thermal Strain in Material Extrusion Additive Manufacturing
Increased interest in and use of additive manufacturing has made it an important component of advanced manufacturing in the last decade. Material Extrusion Additive Manufacturing (MatEx) has seen a shift from a rapid prototyping method harnessed only in parts of industry due to machine costs, to something widely available and employed at the consumer level, for hobbyists and craftspeople, and industrial level, because falling machine costs have simplified investment decisions. At the same time MatEx systems have been scaled up in size from desktop scale Fused Filament Fabrication (FFF) systems to room scale Big Area Additive Manufacturing (BAAM). Today MatEx is still used for rapid prototyping, but it has also found application in molds for fiber layup processes up to the scale of wind turbine blades. Despite this expansion in interest and use, MatEx continues to be held back by poor part performance, relative to more traditional methods such as injection molding, and lack of reliability and user expertise. In this dissertation, a previously unreported phenomenon, irreversible thermal strain (ITε), is described and explored. Understanding ITε improves our understanding of MatEx and allows for tighter dimensional control of parts over time (each of which speaks to extant challenges in MatEx adoption). It was found that ITε occurs in multiple materials: ABS, an amorphous polymer, and PLA, a semi-crystalline one, suggesting a number of polymers may exhibit it. Control over ITε was achieved by tying its magnitude back to part layer thickness and its directionality to the direction of roads within parts. This was explained in a detail by a micromechanical model for MatEx described in this document. The model also allows for better description of stress-strain response in MatEx parts broadly. Expanding MatEx into new areas, one-way shape memory in a commodity thermoplastic, ABS, was shown. Thermal history of polymers heavily influences their performance and MatEx thermal histories are difficult to measure experimentally. To this end, a finite element model of heat transfer in the part during a MatEx build was developed and validated against experimental data for a simple geometry. The application of the model to more complex geometries was also shown. Print speed was predicted to have little impact on bonds within parts, consistent with work in the literature. Thermal diffusivity was also predicted to have a small impact, though larger than print speed. Comparisons of FFF and BAAM demonstrated that, while the processes are similar, the size scale difference changes how they respond to process parameter and material property changes, such as print speed or thermal diffusivity, with FFF having a larger response to thermal diffusivity and a smaller response to print speed. From this experimental and simulation work, understanding of MatEx has been improved. New applications have been shown and rational design of both MatEx processes and materials for MatEx has been enabled.